Final Report REVIEW OF MITIGATION MEASURES … · Review of Mitigation Measures Used To Deal With Habitat Fragmentation By Major Linear Infrastructure : Report: March 2009 4 There

December 2008

Final Report

REVIEW OF MITIGATION MEASURES USED TO DEAL WITH THE ISSUES OF HABITAT FRAGMENTATION

Authors & Title: van der Ree, R., Clarkson, D.T., Holland, K., Gulle, N., Budden M., 2008. Review of Mitigation Measures used to deal with the Issue of Habitat Fragmentation by Major Linear Infrastructure, Report for Department of Environment, Water, Heritage and the Arts (DEWHA), Contract No. 025/2006, Published by DEWHA.

Acknowledgements: The authors would like to acknowledge the following for their advice and comments in the preparation of this report: Kelly Benson, Sylvana Maas, Vanessa Place, Chris Murphy, Joel Benjamin, Dianne Cameron, Miriam Goosem, Rhidian Harrington, Scott Watson, Leanne Maas, James Leslie, Kevin Roberts, Brendan Taylor, Tony Mitchell, Ian Harris, Mark Fitzgerald, Jay Quadrio and Trevor Ferris.

Photo acknowledgements: Australian Research Centre for Urban Ecology (van der Re, R.) SMEC (Ferris. T.), Thiess Pty Ltd (Bax, D).

Copyright: The report was prepared by SMEC (Australia) and the Australian Research Centre for Urban Ecology for Department of Environment and Water Resources. The concepts and information contained in this document are the property of Australian Government Department of Environment, Water, Heritage and the Arts.

Disclaimer: The views and opinions expressed in this report do not necessarily reflect those of the Commonwealth Government. While reasonable efforts have been made to ensure that the contents of this publication are factually correct the Commonwealth Government does not accept responsibility for the accuracy or completeness of the contents, and shall not be liable for any loss or damage that may be occasioned directly or indirectly through the use of, or reliance on, the report. Readers should exercise their own skill and care with respect to their use of the material published in this report and that users carefully evaluate the accuracy, currency, completeness and relevance of the material for their purposes.

Review of Mitigation Measures Used To Deal With Habitat Fragmentation By Major Linear Infrastructure : Report : March 2009 1

CO

NT

EN

TS

Table of Contents

1 Executive Summary ..........................................................3

2 Terms of Reference...........................................................6

3 Effects of Linear Infrastructure on Wildlife.......................7

3.1 Introduction .................................................................................. 7

3.2 Direct and Indirect Loss of Habitat ................................................... 8

3.3 Effects on Wildlife .......................................................................... 9

4 Effectiveness of Measures to Mitigate Habitat Fragmentation – A Literature Review ..................................................... 12

4.1 Introduction .................................................................................12

4.2 Methods ......................................................................................13

4.3 Results ........................................................................................20

4.4 Discussion....................................................................................20

5 Cost – Benefit Analysis of Habitat Fragmentation Mitigation Measures ........................................................................ 20

5.1 Paucity of Hard Data for Effective Financial Cost and Ecological Benefit20

5.2 Variables in the Cost – Benefit Analysis ...........................................20

5.3 Accounting for the Differences in Cost .............................................20

6 General Principles........................................................... 20

6.1 The Principles ...............................................................................20

7 Conclusions and Recommendations................................ 20

7.1 Effects from major linear infrastructure ...........................................20

7.2 The need for crossing structures .....................................................20

7.3 The use of wildlife crossing structures .............................................20

7.4 The efficacy of wildlife crossing structures........................................20

7.5 Pre-Construction Studies ...............................................................20

7.6 Monitoring and evaluation..............................................................20

7.7 Information dissemination..............................................................20

7.8 Cost and benefit ...........................................................................20


7.9 General principles and national guidelines........................................20

7.10 Standard definition of terms...........................................................20

7.11 Future research ............................................................................20

8 Summaries of Relevant Literature .................................. 20

8.1 Australian Refereed Journal and Book Articles ..................................20

8.2 Australian Reports, Theses and Conference Proceedings ....................20

8.3 Australian ‘Best Practice’ manuals and Guidelines .............................20

8.4 International refereed journal publications .......................................20

8.5 International Reports, Theses and Conference Proceedings ................20

8.6 International Reviews and Syntheses ..............................................20

8.7 International ‘Best Practice’ Manuals and Guidelines..........................20

9 References...................................................................... 20

APPENDIX 1 : Summary of Criteria Against which each Publication was Assessed

APPENDIX 2 : Miscellaneous Engineering Drawings


EX

EC

UT

IVE

SU

MM

AR

Y 1 Executive Summary

Ecological crossing structures on major linear infrastructure do work to mitigate habitat fragmentation. The crucial questions are how well certain measures work, in what circumstances, which structures work best, for which species, and how performance can be improved.

This report identifies and reviews an extensive body of literature on the effectiveness of measures that aim to mitigate the fragmentation effects of linear infrastructure, such as roads, railway lines, and utility easements (e.g. powerlines, pipelines). A broad purpose of this report is to provide information on effectiveness to assist the Department of Environment, Water, Heritage and the Arts evaluate proposals submitted to it under the Environment Protection and Biodiversity Act (1999).

It was a review of all key documents in Australia, including information published within scientific refereed journals, conference proceedings and consultants reports that provide data on the use of mitigation measures. Documents from overseas were included only if published in a refereed scientific journal or where it demonstrated a high standard of monitoring and of significance to Australia.

The general effects of linear infrastructure and traffic on the environment are numerous and varied, hence these overall effects are only mentioned briefly in the introduction section of this report. This report aims to identify the types of mitigation measures that have been installed, the species that use them and to evaluate the extent to which they restore connectivity.

Over 50 scientific articles and consultant’s reports were identified as possibly providing sufficient information on the type of mitigation measure, and the type and rate of use by different species or faunal groups. Twenty-nine publications described efforts in Australia to facilitate the safe crossing of linear infrastructure by wildlife.

Fragmentation of habitat has been mitigated by the installation of wildlife-specific structures (e.g. culverts, underpasses, overpasses, glider poles and canopy bridges), the use of existing drainage structures (culverts, pipes and bridges) and the design of linear infrastructure to maintain connectivity (e.g. maintaining canopy connection above the road or in gullies below powerlines, or elevating the linear infrastructure above the vegetation).


There is a great deal of ambiguity in the literature on the definition of structure type. The greatest confusion appears to be in the case of culverts, underpasses, tunnels and bridges. In these four situations, animals go under the linear infrastructure; henceforth, we propose that the structure be named by its design (e.g. 3m x 3m box-cell culvert), rather than its function (e.g. underpass).

Most mitigation measures facilitated the crossing of the linear infrastructure by a variety of species of wildlife. Crossing structures and other mitigation measures were successful at reducing the fragmentation effects of linear infrastructure for a wide range of species, particularly mammals. Different monitoring methods were used to document the use of structures by wildlife. Complete traverses were documented in a limited number of cases, while most inferred complete crossings by indirect evidence (e.g. footprints).

The stated or implied goal of most mitigation projects was to ‘‘restore connectivity’’. However it is difficult to assess effectiveness without a specific and measurable goal. Often the goal is not stated, is ambiguous or is imprecise. Thus, for example, it is not clear whether the goal of some of the mitigation measures were to restore connectivity to a ‘pre-road’ condition, a ‘pre-road upgrade’ condition or to some other pre-determined level that will maintain population processes.

There is sufficient evidence to demonstrate that many species of terrestrial vertebrates will use a range of crossing structures. The literature to date suggests that mammals are more likely to cross than other species. This is most likely an artefact of sampling methods, as they are targeted towards medium/large mammals. There is very little information on the use of mitigation structures by invertebrates, many of which are likely to be greatly affected by fragmentation effects of linear infrastructures.

In summary, many species of wildlife will use underpasses and/or overpasses to cross linear infrastructure, including devices not specifically designed for the passage of wildlife. The distribution and abundance of wildlife, as well as a range of landscape, habitat and structural factors, have been shown to influence the rate of use.

Many important and fundamental research questions remain unanswered, both within Australia and overseas. In the light of this we would advocate focusing on fewer, higher order research questions.


After conducting this review we identified areas needing attention and make some recommendations which are contained in Section 7 of this report. In summary some of these are that:

� Preliminary ecological studies of the potential impact of possible new major linear infrastructure should be conducted at the earliest possible moment to give the best opportunity to design and conduct an adequate and useful base line study. These preliminary ecological studies should be properly designed, and adequately funded.

� Greater use be made of the SMART approach to evaluate mitigation effectiveness (this requires that each goal is specific, measurable, achievable, realistic and timeframed).

� Monitoring be an integral part of the construction and management process for all major infrastructure projects.

� General principles be sent to all parties involved in the funding, commissioning, reviewing, specifying, designing and constructing of major linear infrastructure. These general principles are:

� fragmentation is only one of the effects of linear infrastructure

� avoid environmentally sensitive areas

� identify the nature of the issues

� better to connect than fragment

� identify the goals for mitigation (use the SMART technique)

� design mitigation structures for faunal groups, communities and ecosystem processes

� mitigation structures should be for a wide range of species

� understand conditions and populations adjacent to structures

� conduct and support targeted research, and

� monitoring should be an integral part of the construction and management process.

� The Department of Environment, Water, Heritage and the Arts take a lead in developing national guidelines based on the general principles.


MIT

IGA

TIO

N M

EA

SU

RE

S:

A L

ITE

RA

TU

RE

RE

VIE

W 2 Terms of Reference

The objectives (as set by Department of Environment, Water, Heritage and the Arts) for this comprehensive literature review were to assess the:

� effectiveness of mitigation measures employed to ameliorate the habitat fragmentation impacts of major infrastructure;

� past and present monitoring programs of the mitigation effects of major infrastructure, including their scientific merit; and

� cost-benefits of mitigation measures, in relation to overall infrastructure project costs.


3 Effects of Linear Infrastructure on Wildlife

3.1 Introduction

Roads, train lines, powerlines and other linear infrastructure are pervasive components of landscapes throughout the world. There is a growing recognition of their deleterious impacts on the natural environment and the need to quantify and mitigate these impacts (Andrews 1990; Spellerberg 1998; Forman et al. 2002; Donaldson and Bennett 2004; Goosem 2004; Davenport and Davenport 2006; Roedenbeck et al. 2007). The effects of linear infrastructure, and the traffic along them, are diverse and include many direct and indirect effects, such as:

� loss, fragmentation and degradation of habitat

� incursion of weeds, disease, dust, pollution and feral animals

� direct mortality of wildlife, due to collision with vehicles

� disruption of movements due to the creation of barriers

� altered microclimatic conditions, and

� disturbance due to vehicle movement, vehicle noise, headlights, and other light sources.

The type and severity of each effect, and the distance it extends into adjacent land varies according to the attributes of the road and surrounding landscape.

The ‘road-effect zone’ is defined as the area over which significant ecological effects of a road and its traffic extend into the adjacent landscape (Forman and Deblinger, 2000). Depending on the width of the road, traffic volume, vehicle speed and characteristics of the adjacent landscape (e.g. slope, vegetation type, direction and speed of prevailing winds), these effects can be observed for distances that may exceed 1000 metres for arterial roads with large volumes of traffic (see Forman et al. 2002 for summary of road effects and distances).

Much of the research quantifying the effects of roads and traffic on the environment has been conducted in North America and Europe, and along arterial roads that carry large volumes of traffic. There is limited direct information on these effects on Australian species and landscapes, or on narrow roads that have relatively low traffic volumes.


The extent of information on the effects of powerlines and other linear infrastructure on the natural environment is small compared to the study of road effects, but there is a growing body of knowledge in Australia (e.g. Goldingay and Whelan 1997; Goosem 1997; Goosem and Marsh 1997; Goosem 2004; Clarke et al. 2006) However, it is believed the general principles learned from the work in North America and Europe can be used to make inferences on the possible effects on Australian species, communities and landscapes. In addition, there are now an increasing number of studies on the specific effects of roads and other linear infrastructure on the Australian environment. Hence, what follows is a brief summary of some of the major effects of linear infrastructure and traffic on the natural environment. Readers interested in a more detailed summary and synthesis of the effects of linear infrastructure on the environment are referred to the publications given at the start of Section 4.1.

3.2 Direct and Indirect Loss of Habitat

3.2.1 Effects on Vegetation

Much natural vegetation along linear infrastructure corridors is lost during construction and in post-construction maintenance activities. To prevent and ameliorate this loss and degradation of vegetation, linear infrastructure should be:

� aligned to minimise the direct loss of habitat

� avoid rare or threatened species or communities, and

� protect valuable habitat components, such as large hollow-bearing trees.

In locations where the alignment cannot be altered, options for ‘offsetting’ or compensating for the clearing of vegetation at the road construction sites should be taken. Currently there is considerable discussion about the merit and efficacy of the different compensation arrangements (Cuperus et al. 1999; Gibbons and Lindenmayer 2007).

There may be ongoing indirect loss of habitat adjacent to the linear infrastructure that occurs as a result of incremental removal or degradation of habitat. Altered hydrological, nutrient and microclimatic conditions along a linear clearing (e.g. Pohlman et al. 2007) may result in the growth of plant species that prefer more sunlight and water (e.g. Lee 2006), increasing growth rates and fecundity (e.g. Lamont et al.


1994a; Lamont et al. 1994b). These novel conditions are made available through increased water run-off from the road surface, opening up of the canopy and regular disturbance regimes (Spooner, 2005). Increased productivity of plants and especially of weedy species adjacent to the linear clearing is a commonly reported effect when the clearing traverses relatively natural areas. A study of the effects of roads and traffic on the plant species composition of heath-land vegetation in Hampshire UK, found that there was enhanced growth of heather (Calluna vulgaris) and grass species (Mollinia caerulea) near the road than further away (Angold 1997). This effect extended for up to 200 metres adjacent to a dual carriageway highway carrying approximately 35 000 vehicles during a 12-hour period. Data from nine sites adjacent to narrower roads with traffic volume ranging from 800 – 10 000 vehicles per 12 hrs indicates the edge effect is predicted to be 15 metres for 100 vehicles per 12 hrs, and 25 metres for 300 – 350 vehicles per 12 hours (Angold 1997). Vegetation adjacent to roads may also experience higher rates of defoliation due to an increase in the density of herbivorous insects compared to further away. Increased levels of foliar nitrogen through pollution adjacent to the road could also be a primary mechanism supporting higher insect densities and higher rates of herbivory activity (Port and Thompson 1980; Angold 1997). These factors are also assumed to be occurring in Australia.

3.3 Effects on Wildlife

3.3.1 Loss of Habitat

Direct and indirect loss of habitat caused by linear infrastructure will also affect fauna. Loss of habitat will decrease the viability of populations by reducing the size of the population that can be supported in that area. Quantifying the amount of habitat lost directly is relatively self-evident, while indirect losses are more difficult to quantify. A study of the Horned Lark in Illinois, USA demonstrated lower population densities in farm paddocks within 200 metres of country roads with 300 – 3000 vehicles per day (vpd) compared with further away (Clark and Karr 1979). A similar effect was evident near Boston, Massachusetts USA, where the presence and breeding of grassland birds adjacent to a two-lane highway (15 000 – 30 000 vpd) was reduced for up to 700 metres, and up to 400 metres when 8000 –

15 000 vpd was present (Forman et al. 2002a). The same study found


that traffic flows of 3000 – 8000 vpd had no obvious effect on the presence or breeding of grassland birds (Forman et al. 2002a). In contrast, the abundance and richness of forest birds was reduced by 20 per cent adjacent to a powerline easement in southern New South Wales (Baker et al. 1998). Significant reductions in both abundance and richness of this bird community were evident up to 125 metres from edge of the easement (Baker et al. 1998). Similarly, Spotted Owls nesting within 400 metres of a logging road (with an estimated 5 – 50 vpd) had higher levels of stress hormones than owls further away (Wasser et al. 1997).

3.3.1 Habitat Fragmentation

The internal fragmentation of natural areas is recognised as a significant threat in parks and nature reserves. Some park agencies in Australia have programs to close and rehabilitate roads that are no longer required (Donaldson and Bennett 2004).

The division of habitat into smaller fragments results in lower population sizes. When roads act as a complete barrier or selective filter to movement, as is the case for many wildlife species (Forman et al. 2002b), these smaller populations may not be connected to other populations, and hence they are at a higher risk of extinction. This is because new individuals or plant propagules are unable to supplement a declining population or to re-establish a locally extinct population. The main factors influencing the barrier effect of a road relate to:

� road width

� traffic volume, and

� behaviour of the species.

Species of animals most at risk of population fragmentation due to roads and traffic include species that are unwilling to travel across cleared areas (Forman et al. 2002b). Birds that favour large blocks of habitat, such as ‘forest-interior’ species, are more likely to be affected than ‘edge species’ that can feed in open areas but also occupy forest. Studies of the movements of understorey birds in the Amazonian rainforest found that roads with a 10 – 30 metres wide canopy opening typically formed a territorial boundary while birds more frequently crossed a road where the canopy remained intact (Develey and Stouffer 2001).

Many species of invertebrates and amphibians may also be at risk of


population fragmentation because of low levels of mobility, avoidance of the unsuitable surface of roads and the relatively high potential for collision with vehicles (Mader 1984; Baur and Baur 1990; Reh and Seiz 1990; Gibbs 1998). Some species of reptile are at risk because they avoid roads, as was demonstrated for Blue-tongued Lizards in the suburbs of Sydney, where home range boundaries were aligned with roads, and they actively avoided crossing the roads (Koenig et al. 2001).

Even narrow roads (e.g. 3 metres in width) appeared to inhibit the movement of small mammals (Barnett et al. 1978; Swihart and Slade 1984), but did not completely eliminate road crossing. The response of species is often specific and even species within the same guild (e.g. small terrestrial mammals) have been shown to display different road-crossing abilities (Goosem 2001).

3.3.1 Mortality Due to Collision with Vehicles

Mortality of wildlife due to collision with vehicles is related to the behavioural characteristics of the species, traffic volume and vehicle speed (Dhindsa et al. 1988). Numerous studies have clearly demonstrated that large numbers of creatures in a wide array of species are regularly and frequently killed along roads (Vestjens 1973; Davies et al. 1987; Coulson 1989; Rosen and Lowe 1994; Goosem 1997; Statham and Statham 1997; Scott et al. 1999; Haxton 2000; Hels and Buchwald 2001; Shuttleworth 2001; Taylor and Goldingay 2003). Road mortality has been identified as a major factor contributing to the decline of many species (Fahrig et al. 1995; Vos and Chardon 1998; Jones 2000; Hels and Buchwald 2001). Road mortality may even have significant impacts on population viability for relatively common species, such as the Swamp Wallaby Wallabia bicolor in peri-urban areas (Ramp and Ben-Ami 2006) It is important to note that in many situations, road mortality is not the only factor contributing to population decline. Other factors may include the loss, fragmentation and degradation of habitat, predation pressure, and competition. The rate of mortality is also related to the position of the road within the landscape. Brown et al. (1986) found a higher incidence of bird-mortality where roads and watercourses intersect.

The potential long-term and population-level effects of increased rates of wildlife mortality on local populations are unknown and probably complex. If mortality is a simple linear function of road length and


traffic volume then doubling the number of vehicles and the amount of road surface could simplistically lead to a four-fold increase in the number of individuals killed (e.g. from 10 per year to 40 per year). However, these relationships are unlikely to be linear because each road death will reduce overall population density, potentially reducing the likelihood of other individuals of the same species being killed. The number of species that potentially scavenge on road-killed carcasses can be expected to be drawn to the greater feeding opportunities, thus increasing the probability that scavengers also being killed. Similarly, species that graze on mown vegetation on the road verge are also at higher risk of mortality. A high rate of mortality of Swamp Wallabies on North Stradbroke Island in Queensland occurred because they were attracted to feed on the road verges that were regularly mown (Osawa 1989).

3.3.1 Incursion of Feral Species

Linear infrastructure can have a detrimental impact on wildlife through increased access and rates of predation by feral predators (May and Norton 1996) and Cane Toads (Seabrook and Dettmann 1996; Brown et al. 2006), as well as predation on bird eggs and young still in the nest (Donaldson and Bennett 2004). Increasing the width of the clearings associated with road and powerlines may potentially attract nest predators and lead to a decline in the abundance of some species of bird.

Roads and traffic may also facilitate the invasion of weeds and exotic plants as seeds attached to undercarriages in mud and dirt (Amor and Stevens 1976; Lonsdale and Lane 1994). Vehicles may bring seeds from a large potential catchment and move them across the landscape rapidly. Roadside verges (during and post-construction) are often modified and become more suitable for seed germination.

4 Effectiveness of Measures to Mitigate Habitat Fragmentation – A Literature Review

4.1 Introduction

The objective of this section of the report is to review the effectiveness of measures used to mitigate the fragmentation of habitat by major linear infrastructure.


The fragmentation effect of linear infrastructure occurs when the rate at which animals are able to traverse the developed area is reduced. If all movement is halted, the infrastructure becomes a complete barrier for the species. If a proportion of individuals are capable of crossing it, then it can be termed a semi-permeable barrier or filter. Filter effects may be species, sex or age-specific, or it may affect individuals within the population at random (e.g. van der Ree 2006). Barrier effects may occur due to avoidance of the novel linear infrastructure, or due to increased rates of mortality as crossing is attempted. Avoidance occurs when animals avoid the linear infrastructure due to the establishment of unsuitable habitat or because of the road effect zone (see section 3.1). The avoidance zone extends for a variable distance on either side of the infrastructure and is caused by changes in the habitat, noise or other disturbance. Increased rates of mortality may occur due to direct collision with vehicles (in the case of roads and railway lines) or increased predation as a result of decreased shelter or protection or increased densities of predators.

4.2 Methods

4.2.1 Definitions

According to the Oxford English Reference Dictionary, the definition of mitigation is ‘to make milder or less intense or severe’. Our review focuses in part on the extent to which the barrier or filter effect has been made less severe.

Linear infrastructure is loosely defined as any linear landscape element that has been constructed or modified by humans to allow the passage of people or resources. Linear infrastructure includes:

� roads

� railway lines

� pipelines, and

� powerlines and other utility easements.

The most common linear infrastructure worldwide are roads, and most of the literature we encountered focuses on the effects of roads and traffic and the use of mitigation structures across them. Therefore, ‘road’ and ‘linear infrastructure’ are often used interchangeably throughout this report. Where specific to other types of linear infrastructure (e.g. powerlines or railways) they are mentioned


specifically.

A wildlife crossing structure is defined in this report as ‘a physical structure that increases the permeability of the road or other linear infrastructure by facilitating the safe passage of animals over or under it and in the case of roads and railways, preventing collision with vehicles’. Wildlife crossing structures may be purpose built for wildlife or may primarily serve other functions (e.g. water drainage or access by humans). Other methods to mitigate the barrier effect of linear infrastructure include systems that are not strictly ‘wildlife crossing structures’ (Goosem 2004). These include maintaining canopy connectivity above the road or below a bridge, installing powerlines above the canopy of the forest with clearing of vegetation just for pylons, or establishing transverse strips of vegetation at low points (e.g. gullies and drainage lines) across powerline clearings (Goosem 2004).

In the literature there is considerable confusion and interchangeable use of terms when describing mitigation structures. We have developed the following terms and definitions (Table 1) to reduce confusion and provide consistency when describing the mitigation structures. In essence, we propose that a mitigation measure be described according to its specific structure, rather than its intended use. We propose to use the terms ‘underpass’ and ‘overpass’ as general terms that describe a collection of structures.

In this report we have classified and reviewed all structures in the studies according to their dimension and form, irrespective of the names given them by the authors. When we were unable to classify mitigation structures into one of our specific categories, they are referred to by the general term of ‘underpass’ or ‘overpass’.

Table 1: Definition of engineering options to mitigate the fragmentation effects of linear infrastructure

Title Description

OVERPASS* Allows passage of animals above the road

Land bridge Also known as eco-duct or wildlife bridge. This is a

(typically) wide (30 – 70 metres) bridge that extends over

the road. The bridge has soil on it, and is planted with

vegetation and enhanced with other habitat features (e.g.


logs, rocks, water-body etc) (Figs. 1 – 3)

Overpass (small roads) This bridge above the major linear infrastructure is typically

to allow human access across the road. This overpass is

typically narrow and not hourglass shaped. The road on the

overpass is typically a minor road – it may be unsealed,

single lane etc

Canopy bridge This is a rope or pole suspended above the traffic, either

from vertical poles or from trees. Typically installed for

arboreal and scansorial species (Figs. 8 – 10)

Glider pole These are vertical poles placed in the centre median or on

the road verge, and provide species that glide intermediate

landing and launch opportunities (Fig. 11)

Local traffic

management

Devices to reduce the speed or volume of traffic – e.g. road

closures, chicanes, crosswalks, lighting, signage (Fig 14)

UNDERPASS* Allows the passage of animals below the major linear

infrastructure

Culvert Culverts are typically square, rectangular or half-circle in

shape and may be purpose built for fauna passage or water

drainage, or a combination of both. They are typically pre-

cast concrete cells or arches made of steel (Figs 4 – 6). By

definition, culverts were originally used to carry water.

However, engineers and road designers are familiar with the

size and shape of culverts, and hence we suggest the

continued use of the term culvert to describe this type of

underpass.

Tunnel Tunnels are typically round pipes of relatively small

diameter (e.g. < 1.5 metres diameter). May also be termed

‘eco-pipe’.


Bridge

A bridge is a structure that maintains the grade of the road

or elevates the traffic above the surrounding land, allowing

animals the opportunity to pass under the road. When used

to mitigate the barrier effect of linear infrastructure, the

primary function is often to facilitate water drainage or the

movement of local human traffic, and secondarily to

facilitate the passage of wildlife (Fig. 7)

NON-STRUCTURAL

MITIGATION

This type of mitigation allows for sensitive road designs that

facilitate ‘natural’ permeability

Canopy connectivity

The width of the linear clearing is kept sufficiently small to

allow the tree canopy to remain continuous above the

clearing, or where not continuous, sufficiently small to allow

gliders (and other volant species) to safely traverse the

clearing

At-grade crossings

Vegetation or other habitat features (e.g. rocks, fallen

timber) are strategically planted or allowed to regrow such

that animals are directed to preferred crossing locations

where they are required to cross the linear infrastructure

without the aid of any structures (i.e. similar to a pedestrian

crossing)

Elevating the linear

infrastructure

The road or powerline is elevated above the vegetation to

minimise clearing (clearing only required for bridge piers or

pylons) and allow natural vegetation to grow under the

infrastructure

Corridor plantings

Are strips of vegetation, similar to that on either side of the

linear clearing that traverse the clearing and provide

corridors for animal movement.

* There was considerable overlap in use of terms to describe crossing structures,

particularly for underpasses. The definitions in this table are an attempt to reflect

their design and method of construction, rather than their potential use.


Figure 1: Is an example of a land bridges or eco-duct overpass land bridge

nearing construction along the Pacific Highway in NE NSW with a front

elevation.


Figures 2 and 3: are two operational land bridges in the Netherlands

Figure 4-6: Are examples of culverts facilitating the movement of wildlife

under linear infrastructure. The top two photos are culverts under the East

Evelyn Rd in the Atherton Tablelands, and the lower photo is a typical box-

cell culvert, on this occasion under a railway line.


Figure 7: Example of a

bridge underpass, at

Slaty Creek, Calder

Freeway, Central

Victoria

Figures 8-10: Are examples of canopy bridges, including Wakehurst Parkway

near Sydney (top left), Atherton Tablelands (bottom left) and across the

Pacific Highway NE NSW (right).


Figure 11: Above right is the first experimental glider pole, located at Termeil

along the Princes Highway in SE NSW. Figures 12-13:Top left is the

completed Karuah aerial fauna crossing and bottom left is the crossing being

used at night (David Box, 2006)


Figure 14: An example of local area traffic management or traffic calming

devices, such as signage, speed humps and crosswalks.

Mitigation structures are designed with the needs of the infrastructure and topography very much in mind. Therefore the particular designs of mitigation structures vary. The variation between similar functional structures between different projects can vary considerably. Just a small sample of some engineering drawings is included by way of example at Appendix 2.

4.2.1 Literature Sources

The information for this critical review was sourced from an extensive literature search. It was obtained by searching the ISI Web of Science database in March 2007 using the terms ‘underpass, overpass, culvert, tunnel and barrier’ in combination with the terms ‘road, powerline, wildlife and fauna’. Extensive reference lists from a number of general ‘road ecology’ reviews (Bennett 1990; 1999; Forman et al. 2002b; Davenport and Davenport 2006) were searched. References from each article obtained were also searched. Conference presentations were only included if the proceedings were available (i.e. PowerPoint files of the actual presentation excluded). Only studies that presented new data on the use or effectiveness of wildlife crossing structures were included in the critical review. Papers that summarised published data, gave an overview of projects within their jurisdiction, or discussed projects that were in the planning or construction stages were excluded from the


review because they did not explicitly provide information on the use or effectiveness of mitigation measures. Furthermore, we have found that the mitigation options proposed in reports are often very different to the final structure or system that is finally constructed. In the case of summaries of existing data we aimed to obtain primary data sources to ensure that that misquoting of unsighted studies did not occur. Studies that included other aspects of road ecology and wildlife (e.g. rate or location of road-kill, effectiveness of exclusion fencing, animal behaviour or survival after translocation, pre-road upgrade conditions) were only included if they also included data on use of mitigation to restore connectivity. Our aim was to locate and critically review the most reliable and scientifically defensible data we could find. Reviews or syntheses that we considered important or authoritative were summarised (e.g. Little et al. 2002). The papers that were included in the review were summarised according to the criteria given in Appendix 1

The literature found during these searches was assessed against the following criteria to determine if they were to be included in the review:

1. Does the publication provide information on the need for or design of structures or other systems to mitigate the barrier effect of major linear infrastructure? Does the publication provide information, data or subjective evaluation of the usage or effectiveness of mitigation structures or systems?

2. Did the development impact upon any matter of national environmental significance (e.g. were any species or communities listed under the EPBC Act affected by the linear infrastructure)?

3. Did the development impact upon any matter of state environmental significance (e.g. were any species or communities listed under state legislation affected by the linear infrastructure)?

4. If the publication contains information from international examples or studies can it be translated or applied to the Australian situation? In other words, can the overseas situation provide analogous biological or ecological information? Can it be used to provide a standard or framework for monitoring the effectiveness of the mitigation device? Do the international sources ‘fill an important gap’ in the Australian literature/understanding?


4.2.1 Referrals under the EPBC Act (1999)

Considerable effort and time was spent searching through the EPBC referral list available on the Department of Environment, Water, Heritage and the Arts website, looking for relevant referrals that involved mitigation measures in the form of fauna movement structures associated with major linear infrastructure. After completing this search a summary list was provided to the Departmental contact officer informing her of the additional Departmental records we wished to examine, if these could be readily obtained. The types of reports of interest were EIS, Flora and Fauna Management Plans, controlled actions reports and in-house Department of Environment, Water, Heritage and the Arts monitoring reports.

A SMEC employee was invited to the Department of Environment, Water, and the Arts office in Canberra, where they were given access to library databases, to search for relevant information and to apply for access to reports from the library. No further relevant information was found on the databases; beyond what was already discovered from the EPBC referral lists available on the internet. Some reports were only preliminary, because construction work was in progress (such as the Tugun Motorway Bypass Project) or monitoring had not yet commenced.

We noted that all other projects deemed not to be ‘controlled actions’ were passed back to the relevant state departments and agencies for any other action they deemed appropriate. It was not possible to get definitive data on studies state departments or agencies may have initiated for ‘non-controlled actions’.


4.3 Results

4.3.1 Number of Papers and Geographic Location of Study

Fifty-nine studies that fitted our search criteria (see section 5.2.2) for structures used as wildlife crossings by fauna were found and reviewed. These 59 studies included:

� all publications in English-language scientific refereed journals

� all Australian consultants reports

� a small number of conference proceedings, and

� articles published between 1975 and 2007.

These studies were conducted in the following countries:

� one study each in France, Denmark, Brazil, Sweden and Portugal

� two in the Netherlands

� four in Spain

� five in Canada

� fourteen in the United States of America, and

� twenty eight in Australia.

Of the 28 studies from Australia:

� twenty one were consultant’s reports

� two were conference proceedings, and

� five were from refereed scientific journals (2 published, and 2 currently in review or in preparation).

The majority of crossing structures in Australia have been installed along the Pacific Highway in north-east New South Wales (Australian Museum Business Services 2001d, 2001g, 2001f, 2001b, 2001c, 2001a, 2002b, 2002a; Fitzgerald 2003; Taylor and Goldingay 2003; Fitzgerald 2004, 2005; Hayes 2006), with a smaller number in Victoria (Mansergh and Scotts 1989; Abson and Lawrence 2003a, 2003b; Abson 2004; Hamer 2006), Queensland (Jones et al. 2004; Goosem 2005) and Western Australia (Ecologia Environmental Consultants 1995). Internationally, mitigation structures are widespread across the USA (see Irwin et al. 2003; 2005 for examples), and Europe, with a notable concentration in Banff National Park in Canada (Clevenger and Waltho


1999, 2000; Clevenger et al. 2003; Clevenger and Waltho 2003; Clevenger and Waltho 2005).

4.3.1 Number and Type of Structures and Use by Australian Wildlife

The majority of mitigation structures installed in Australia and around the world are underpasses, and specifically culverts. Of the 527 structures that were reported on in the 59 studies we reviewed, 88 per cent (n = 462) allowed animals to move under the linear infrastructure and the remainder allowed passage above the road. The underpasses included:

� culverts (339 examples)

� bridges (71)

� underpasses of unknown type (43), and

� tunnels (10).

Overpasses included:

� land bridges (20)

� overpasses with small roads (18)

� canopy bridges (8)

� glider poles (1), and

� other devices such as crosswalks, and signage (19).

Within Australia, wildlife crossing structures have been monitored for use in eastern and south-western Australia (Table 2). A conservative estimate of the total number of structures within Australia that facilitate the crossing of roads by wildlife that have been formally evaluated and reported on is approximately 100 (Table 2). Many of the structures could well have been included in a number of different investigations and reported by different authors (e.g. series of studies along the Pacific Highway in NE New South Wales).

Internationally, the mean number of crossing structures investigated per study was 9.9 (range 1 – 82), with 23 of the studies focusing on one (n = 12), two (n = 3) or three (n = 5) structures. Within Australia, the mean number of mitigation structures per study was 6.5 (range 1 – 22), with 11 studies focusing on one (n = 4), two (n = 2) or three (n =


5) structures.

Table 2: Location and type of structure across roads designed to mitigate fragmentation effects in Australia

Road Location Target

Species

Number and

Type of

Structures

Status (as at

March 2007)

Pacific Highway, NE NSW – Herons Creek

Wildlife 1 x bridge underpass 1 x box culvert 1 x pipe culvert

Operational Operational Operational

Pacific Highway, NE NSW – Brunswick Heads

Wildlife 2 x bridge underpass 11 x box culvert

Operational Operational

Pacific Highway, NE NSW – Buladelah – Coolongolook

Wildlife 2 x multi-span bridge 17 x box culvert 3 x concrete pre-cast arch


Pacific Highway, NE NSW – Taree

Wildlife 2 x bridge underpass 1 x steel arch 4 x fauna / drainage underpass


Pacific Highway, NE NSW – Yelgun – Chinderah

Wildlife 7 x bridge underpass 3 x box culvert 2 x steel arch 2 x land-bridge overpass

Operational Operational Operational Operational

Pacific Highway, NE NSW – Karuah Bypass

Squirrel Glider, arboreal marsupials

5 x rope-ladder canopy bridges

Operational

Calder Freeway, Slaty Creek (central Vic)

Wildlife 2 x bridge 6 x culverts

Operational Operational

Calder Freeway Kyneton to Faraday (central Vic)

Brush-tailed Phascogale

5 x culvert 2 x canopy rope bridge 9 x at-grade crossings 6 x bridge underpass

Operational

Calder Freeway Harcourt Section (central Vic)

Brush-tailed Phascogale

2 x bridge underpasses

Planning and construction stage


6 x box culvert 3 x at-grade crossings

Goulburn Valley Freeway (central Vic)

Squirrel Glider Canopy bridges Glider poles

Planning stage

Compton Rd (Brisbane, Qld)

Wildlife 2 x box culvert 1 x land bridge overpass 3 x rope-ladder canopy bridges 8 x glider poles

Operational

Atherton Tablelands (Qld) Lumholtz’ Tree-kangaroo, Cassowary, ringtail possums

2 x rope-ladder canopy bridges 4 x Culvert

Operational

Princes Highway, Pakenham Bypass, SE suburbs of Melbourne

Growling Grass Frog

9 x culverts 4 x bridges

Princes Highway (Termeil, southern NSW)

Gliders 1 x experimental glider pole

Operational

Princes Highway (near Bellbird, East Gippsland)

Long-footed Potoroo

3 x round culvert Operational

Wakehurst Parkway, Sydney

Arboreal marsupials

1 x canopy bridge Operational

Lady Game Drive, East Lindfield, Sydney

Arboreal marsupials

2 x canopy bridge Operational

F3 Freeway (Sydney to Newcastle, NSW)

Wildlife 1 x box cell culvert 1 x arch culvert 1 x pipe culvert

Operational

Hume Freeway, Craigieburn Bypass

Growling Grass Frog

6 x box culverts 4 x combined drainage culvert

Operational

Hume Freeway, NE Victoria (Longwood to Benalla)

arboreal marsupials, primarily Squirrel Glider,

2 x rope-ladder canopy bridges 3 x glider pole sites

Operational

Hume Freeway, Albury to Tarcutta

Squirrel Glider, reptiles

Numerous rope-ladder canopy bridges, culverts, glider poles

Planning and construction stage

Kwinana Freeway, Perth Southern Brown Bandicoot

6 x pipe culverts Operational

Roe Highway Stage 7 Southern Brown Bandicoot

3 x box culverts Operational

In Australia 10 species of amphibian, 39 species of bird, three species of


invertebrate, 52 species of mammal and 13 species of reptile have been observed using underpasses or overpasses. Most species have been observed within underpasses. A majority of the records of birds seem to be coming from the bridge underpass at Slaty Creek (Abson and Lawrence 2003b). The relatively recent construction of a small number of overpasses in Australia seems to account for the bulk of reports (three land bridges and five canopy bridges). The recent construction of many structures may also partially account for the comparatively low reported rate of use by wildlife because rate of use is expected to increase due to the establishment of vegetation and habituation over time.

Three species of wildlife listed under the EPBC Act were detected. These were the:

� mountain pygmy-possum, in culverts at Mt Higginbotham (Mansergh and Scotts 1989; van der Ree et al. in review), and

� spot-tailed quoll, observed in culverts and land bridges.

Wildlife crossing structures were installed under the Craigieburn Bypass of the Hume Freeway near Melbourne (Hamer 2006) and three round culverts were installed in November 1986 under the Princes Highway near Bellbird in East Gippsland for the Long-footed Potoroo (Scotts and Seebeck, 1989). Monitoring of one of the culverts using a remotely triggered camera from January to April 1987 and sporadic monitoring in subsequent years failed to detect any long-footed potoroos using the culverts (Scotts and Seebeck 1989, Tony Mitchell DSE, pers comm.). Similarly, no Growling Grass Frogs were detected using their culverts during the first year of surveying (Hamer 2006).


Table 3: List of Australian species (or species group) observed using culverts, tunnels, underpasses, land bridges and canopy bridges / glider pole

Fauna group or

species

(number of species)

Scientific

name

Culvert Tunn

el

Bridge

underpa

ss

Land

bridg

e

Canopy

bridge /

Glider pole

Amphibians (n = 10)

bullfrog, southern

Limnodynastes dumerilii

D

frog or toad spp. b, g-(2,3), h, m-1

g-(1,2,3) g-3, h

frog, brown-striped

Limnodynastes peronii

b, i

frog, leseur's Litoria lesueuri

i

froglet, common Crinia signifera

D

froglet, Victorian smooth

GeocrinaVictoriana

D

*toad, cane Bufo marinus b, g-(1,2,3)

g-(1,2,3) g-(1,3)

toadlet, brown Pseudophryne bibronii

D

tree frog, plains brown

Litoria paraewingi

D

tree frog, Southern brown

Litoria ewingii D

tree frog, whistling Litoria verreauxii verreauxii

D

Birds (n=39)

bird spp. b, g-(1,2,3), h, m-1, o, q

e g-(1,2,3), k, m-1, n-1, o

g-(2,3), h

*blackbird Turdus merula

D

brush turkey Alectura lathami

q

chough, white-winged

Corcorax melanorhamphos

D

cockatoo, gang-gang

Callocephalon fimbriatum

D


cockatoo, sulphur crested

Cacatua galerita

D

cockatoo, yellow-tailed black

Calyptorhynchus funereua

D

crested shrike-tit Falcunculus frontatus

D

currawong, pied Strepera graculina

D

duck spp. g-2 g-2

duck, pacific black Anas superciliosa

D

fairy-wren, superb Malurus cyaneus

D

fantail, grey Rhipidura fuliginosa

D

fantail, rufous Rhipidura rufifrons

D

finch, red-browed Neochmia temporalis

D

flycatcher, restless Myiagra inquieta

D

heron spp. h

heron, white-faced Egretta novaehollandiae

D

honey eater, white eared

Lichenostomus penicillatus

D

honey eater, White napped

Melithreptus albogularis

D

honey eater, yellow-faced

Lichenostomus Chrysops

D

honeyeater, painted Grantiella picta

D

honeyeater, yellow-tuffted

Lichenostomus melanops

D

kookaburra Dacelo novaeguineae

D

magpie Gymnorhina tibicen

D

pigeon, Wonga Leucosarcia melanoleuca

i

quail b, k

raven, Australian Corvus coronoides

D

raven, little Corvus mellori

D

robin, eastern yellow

Eopsaltriaaustralis

m-2 D


rosella, crimson Platycerus elegans

D

scrubwren, white-browed

Sericornis frontalis

D

shrike-trush, grey Colluricincla harmonica

D

silvereye Zosterops lateralis

D

spinebill, eastern Acanthorhnychus tenuirostris

D

swallow, welcome Hirundo neoxena

D

thornbill spp. Acanthiza spp.

D

thornbill, brown Acanthiza apicalis

D

thornbill, buff-rumped

Acanthiza reguloides

D

thornbill, striated Acanthiza lineate

D

tree creeper, white throated

Cormobates leucophaeus

D

wattlebird, red Anthochaera carunculata

D

whistler, golden Pachycephala pectoralis

D

Invertebrates (n=3)

crab, small K

invertebrate sp. b, k

snail k

Mammals (n=52)

antechinus spp. possibly: yellow-footed antechinus, brown antechinus

Antechinusflavipes, Antechinusstuartii

i, m-(1,2,3), n-(1,2,3), o

n-2, o

antechinus, agile Antechinusagilis

D

bandicoot spp. possibly: northern brown, long-nosed

Isoodon macrourus Perameles nasuta

b, g-(1,2,3), h, k, n-(1,2,3), m-(1,2,3), o, q, s

g-(1,2,3), k, m-1, n-(1,2,3), o

g-(1,2,3), h

bandicoot, Long-nosed

Perameles nasuta

c, i


bandicoot, Northern brown

Isoodon macrourus

b, i n-3

#bandicoot, southern brown

Isoodon obesulus

e

bat spp. m-(2,3)

bat, chocolate wattled

Chalinolobus morio

D

bat, eastern freetail Mormopterus spp.

D

bat, Gould's long-eared

Nyctophilus gouldi

D

bat, large forest Vespadelus darlingtoni

D

bat, lesser long-eared

Nyctophilus geoggroyi

D

bat, little forest Vespadelus vulturnus

D

bat, southern forest Vespadelus regulus

D

bat, southern freetail

Mormopterus spp.

D

bettong, rufous Aepyprymus rufescens

m-2

*brown hare Lepus capensis

D

*cat Felis catus b, c, d, g-(1,2,3), h, i, k, m-(1,2,3), q, s

e d, g-(1,2,3), m-1

g-3, s

*cow Bos Taurus g-2 g-1, m-1, n-1

g-2

*dog Canis lupus familiaris

c , d, g-(1,2,3), h, i, m-(1,2,3), n-1, q, s

d, g-(1,2,3), m-1, n-1

g-(1,2,3), H, s

dasyurid (common dunnart, planigale, yellow-footed antechinus)

Sminthopsis murina, Planigale maculata, Antechinusflavipes

s

dunnart spp. possibly common dunnart

possibly Sminthopsis murina

m-(1,3)

eastern false pipistrelle

Falsistrellus tasmaniensis

D


echidna Tachyglossus aculeatus

b, g-(2,3), k, m-(1,2,3), n-3, s

d, g-(1,2)

g-3, s

*fox Vulpes vulpes b, c, d, g-(2,3), h, i, k, m-(1,2,3), n-(1,3)

e d, g-(2,3), k, m-1, n-1

g-2, s

glider, squirrel Petaurus norfolcensis

F, U, v

glider, squirrel/sugar

Petaurus D

glider, sugar Petaurus breviceps

i

hare, brown Lepus capensis

s s

*horse Equus caballus

D g-1

kangaroo, Eastern grey

Macropus giganteus

i, m-2 D s

koala Phascolarctos cinereus

b, g-3, k, n-3

d, k

macropod spp. possibly: Eastern grey kangaroo, swamp wallaby, red-necked wallaby

Macropus giganteus, Wallabia bicolour, Macropus rufogriseus

h, m-(1,2), n-1, s

O H

mammal, small (bush rat, black rat, swamp rat, water rat, yellow-footed antechinus, house mouse, planigale)

Rattus fuscipes, R. rattus, R. lutreolus, Hydromys chrysogaster, Antechinus flavipes, Mus musculus, Planigale maculata.

g-(1,2,3), s

g-(1,2,3) g-3

melomys, Fawn-footed

Melomys cervinipes

Q

mouse spp. b

*mouse, house Musmusculus

h, m-(1,2), n-1, s

e d, g-2, n-1, o

pademelon spp. possibly red-necked

Thylogale spp. possibly T. thetis

m-(1,2,3)


pademelon, Red-legged

Thylogale stigmatica

q

platypus g-(1,2)

possum spp. possibly: brushtail possum or ringtail possum

b, g-(2,3), h, k, m-(1,2), s

d, g-(1,2,3)

g-3, h, s

possum, brushtail spp. possibly mountain or common

Trichosurus caninus or T. vulpecula.

b, d, h, k, m-(1,2,3), n-3, o

d, g-1, k, n-1, o

h F

possum, common brushtail

Trichosurus vulpecular

P, v

possum, common ringtail

Pseudocheirus peregrinus

b D J, V

possum, coppery brushtail

Trichosurus vulpecula johnstoni

q Q

possum, Eastern pygmy

Cercartetus nanus

i

possum, green ringtail

Pseudochirops archeri

r, q

possum, Herbert River ringtail

Pseudochirulusherbertensis

r, q

possum, lemuroid ringtail

Hemibelideus lemuroids

r, q

possum, long-tailed pygmy

Cercartetus caudatus

Q

#possum, mountain pygmy

Burramys parvus

a

possum, striped Dactylopsila trivirgata

Q

potoroo, long-nosed Potorustridactylus

m-3

#quoll, spotted-tailed (a.k.a. tiger quoll)

Dasyurus maculates

i, m-2 g-2

*rabbit Oryctolagus cuniculus

e D

*rabbit/hare Oryctolagus cuniculus orLepus capensis

g-(1,2,3) g-(1,2) g-3

rat spp. possibly: black rat, bush rat, swamp rat or water rat

Rattus rattus, R. fuscipes, R. lutreolus or Hydromys chrysogaster

b, g-(1,2,3), h, k, m-(1,2,3), n-(1,2,3), o

g-(1,2,3), k, m-1, n-(1,2,3), o

g-3, h


*rat, black Rattus rattus b, i, n-2 d, n-(2,3)

rat, bush Rattus fuscipes

b, c, d, i D

rat, swamp Rattus lutreolus

b, i n-2

rat, water Hydromys chrysogaster

k, g-3, m-(2,3), o

K

rodent spp. H, q h

tree-kangaroo, Lumholtz's

Dendrolagus lumholtzi

q Q

wallaby spp. red-necked or swamp

b, k, m-(1,2,3), n-3

K

wallaby, red-necked Macropus rufogriseus

n-3 s

wallaby, swamp Wallabia bicolour

b, c, g-(1,2,3), i, m-(2,3)

d, g-(2,3)

g-(1,2,3), s

wombat Vombatus ursinus

d, i, m-3 D

Reptiles (n=13)

dugite Pseudonaja affinis

e

goanna k, m-(1,2,3)

k, m-1

lizard spp. b, g-3, h, k, m-1, s

K h

lizard, blue-tongued Tiliqua scincoides

m-1 g-3 g-3

shingleback (a.k.a. stumpy-tail, bobtail)

Tiliqua rugosa

e

monitor, lace Varanus varius

b, g-(1,2,3), i, n-(1,3), o

g-(1,2,3), n-(1,3), o

g-(2,3)

monitor, sand Varanus gouldii

e

python spp. possibly diamond python

Morelis spp. possibly M.spilota spilota

i, m-2

skink spp. g-(1,2,3), k, m-1, n-1

g-(1,2,3), m-1, o

g-(1,3), h

skink, Coventry's Niveoscincus coventryii

D


skink, eastern water

Eulamprus quoyii

g-(2,3), i g-(1,2,3)

skink, garden Lampropholis guichenoti

D

skink, McCoys Nannoscincus maccoyi

D

snake spp. b, g-(1,2,3), m-(1,2,3), s

d, g-(1,2,3), n-1

g-(1,3), h

snake, red-bellied black

Pseudechis porphyriacus

m-2

turtle spp. g-3

water dragon, eastern

Physignathusleseuerii

g-(1,2,3), i, m-(1,2,3), n-(1,2,3), o

g-(1,2,3), m-1, n-(1,2,3), o

Letters denote data source (see Table 3 for reference information). Capital letters denote that the structure was ‘preferred’, lower case and not bold denote not preferred, and lower case bold denotes that preference was not determined. For the majority of species and records of use, ‘preference’ was not determined. NB - culverts, tunnels and underpasses are not always clearly defined – hence caution required when interpreting usage for these species. * = introduced species # = species listed as endangered under the EPBC Act 2001

Table 4: Reference details for Table 3 and most commonly detected species using crossing structures

Letter code (as per Table

3)

Author(s) of publication. Refer to reference list for full details

of publication.

Most commonly detected sp (percentage of records)

a Mansergh and Scotts 1989 only reported mountain pygmy possum

b Taylor and Goldingay 2003 bandicoots (25.3%), rats (25.3%)

c Hunt, Dickens and Whelan 1987 rat, dog and fox

d Abson and Lawrence 2003 not given

e Ecologia - Kwinana freeway wildlife underpass study 1995

rabbits (38%)

F Karuah bypass fauna crossing report 2006

brushtail possum (92%)

g-1 Fitzgerald Feb/Mar 2003 cane toad

g-2 Fitzgerald Oct03/Jan2004 swamp wallaby

g-3 Fitzgerald 2005 swamp wallaby

H Hayes 2006 culverts: bandicoot (42 %),


overpass: macropods (42 %)

I AMBS 1997 long-nosed bandicoot (50%), bush rat (27%)

J AMBS Jan 2001 only recorded common ringtail possum

K AMBS Mar 2001 Brunswick Heads bridge underpass: large wallaby (40%). culvert: large wallaby (38%) and bandicoot (32%)

m-1 AMBS Mar 2001 Bulahdelah to Coolongolook

bandicoot

m-2 AMBS July 2001 Bulahdelah to Coolongolook

bandicoot, rat

m-3 AMBS June 2002 Bulahdelah to Coolongolook

bandicoot, goanna, cat

n-1 AMBS Mar 2001 Taree rat, fox

n-2 AMBS June 2001 Taree rat

n-3 AMBS June 2002 Taree culvert: bandicoot, antechinus. bridge underpass: rat

O AMBS Mar 2001 Herons Creek rat

P AMBS June 2006 common brushtail possum (only one record)

Q Goosem, Weston and Bushnell 2005 culverts: bandicoots (24.9%), red-legged pademelon (24.5%), brushtail possum (24.1%). canopy bridge: not given

R Weston 2001 not given

S Bond and Jones 2008 culverts: small mammals: rat/antechinus (34.8 %). land bridge: hare (76.1 %)

T Harris 2007 southern brown bandicoot (other species incl. western grey kangaroo, brushtail possum, rats, skinks, monitor)

U Ball and Goldingay 2008 squirrel glider

V van der Ree et al 2008 common ringtail possum

4.3.1 Type of Linear Infrastructure

Structures to enhance connectivity have been constructed across a number of different types of linear infrastructure. The majority of the 59 studies have focussed on mitigating the effects of:

� roads and traffic (53)

� railway lines (3)

� one on road and railway lines

� one on an oil pipeline, and


� one on a water canal.

We noted it was rare to find that the conditions of the road and/or traffic were fully described. Only eight fully reported the road characteristics and 14 publications fully reported traffic conditions. We defined road characteristics as:

� road width or width of clearing

� number of lanes, and

� presence, width and vegetation characteristics of a median strip.

The significant traffic conditions reported were:

� traffic speed, and

� traffic volume and intensity.

The type of road was typically described in terms of its classification and number of lanes (e.g. ‘major highway’, ‘4-lane interstate highway’, or ‘4-lane divided highway’). This is another potential source of confusion, especially when attempting international comparison. It may be possible to infer road width to some extent from the length of the structure used in mitigation. However there was a pronounced variation in the length of structures, even on the same road or train-line, thus limiting the utility of this approach (e.g. Hunt et al. 1987; Clevenger and Waltho 2000; Ng et al. 2004).

4.3.1 Design, Timing and Duration of Study

The majority of studies measured the rate of use of crossing structures and

came to conclusions about the factors influencing crossing rates. This was

most rigorously achieved by relating the rate of use to habitat, landscape and

physical characteristics of the mitigation structures by using a correlation

and/or regression approach.

A ‘before-after’ comparison approach was evident in 11 of the 59 studies,

which included rates of road-kill before and after mitigation (e.g. Dodd et al.

2004). (Mansergh and Scotts 1989)) conducted an assessment of population

sex ratios and over-winter survival before and after mitigation. Other

researchers assessed the effectiveness of fencing to funnel animals towards

the structures (e.g. Rodriguez et al. 1997; e.g. Cain et al. 2003).

Most studies commenced after the structures had been built and were

therefore unable to include a rigorous assessment of the pre-mitigation

scenario. Similarly, many studies that investigated faunal use of non-wildlife


passages (e.g. drainage culverts) were unable to include a pre-mitigation analysis. A number of studies implied that there were elements of the ‘before and after’ approach, but this was not conclusive or clear from their methods or results. Three sequential studies (Singer 1978; Singer and Doherty 1985; Pedevillano and Wright 1987) reported on the use of the same structures over time, and in combination the three studies provided a before and after approach. One study used a novel approach to test small mammal preferences by translocating animals across the road in the vicinity of different structures, and provided excellent replication by using many animals near multiple structures (McDonald and St Clair 2004). Only one study included a control, and examined the behaviour of mountain goats near two salt licks – one near and the other distant from a highway (Singer 1978).

Most studies were not explicit about the timing of their surveys in relation to structure or road completion. The earliest use of a structure after construction was recorded for the Mountain Pygmy Possum (Burramys parvus) which used a tunnel two weeks after completion (Mansergh and Scotts 1989). Similarly, Golden Lion Tamarins were reported to use a canopy bridge ‘as soon as it was assembled (Valladares Padua et al. 1995).

The duration of monitoring varied across studies, ranging from 3 weeks to 20 years. The mean duration of each reported study in Australia was 1.06 years, which was significantly influenced by a single study which utilised a 20-year census data set (van der Ree et al. in review). Excluding this study, which analysed the 20-year data set, the mean duration of monitoring across the 27 reports, was 4.2 months (range of 3 weeks – 1 year). The mean duration of monitoring of the international studies was 2 years (range 3 months – 6 years). The frequency of monitoring within each study was extremely variable, and included daily (Reed et al. 1975), once per week (Clevenger et al. 2001), and 15 – 22 days per month (Rodriguez et al. 1997). The frequency of monitoring depends in part on the survey technique selected.

4.3.1 Description of Populations of Wildlife and Habitats Adjacent to Roads

Most studies (39 of the 59) gave some description of the vegetation or landform in their study area. However, there was generally insufficient detail given on the vegetation, topography, and altitude at each


structure for the reader to assess habitat suitability for supporting populations of species that may utilise the crossing structures.

Thirty studies incorporated some assessment of the presence or abundance of their target species into their evaluations. The most comprehensive was a calculation of expected crossing rates based on relative animal abundance in adjacent habitats (Clevenger and Waltho 2000; Clevenger et al. 2001).

Examples of various reported methods included:

� radiotracking (e.g. Foster and Humphrey 1995, Australian Museum Business Services 2001e, Cain et al. 2003)

� track or camera counts (e.g. Gloyne and Clevenger 2001; Braden et al. in press), and

� studies that detailed census data of their target species (Mansergh and Scotts 1989; Guyot and Clobert 1997, van der Ree et al. in review).

Literature sources and museum databases were used in 16 studies to evaluate habitat preferences and seasonal fluctuations in abundance. We believe that assessment of animals in adjacent habitat was not relevant in studies that focused primarily on the behaviour of animals using the tunnels (Reed et al. 1975; Singer 1978; Reed 1981; Singer and Doherty 1985; Pedevillano and Wright 1987) or in the single study that used a translocation approach (McDonald and St Clair 2004).

4.3.1 Survey Techniques

A range of techniques was used to identify the use of crossing structures by wildlife:

� the most common technique was tracking pads (33 studies), where a substrate (e.g. sand, soot, ink) was used to record animal footprints, from which the species, direction of travel and number of crossings could be inferred

� twenty studies used video or remotely triggered infra-red still cameras

� other commonly-used techniques included radiotracking (4 studies)

� direct observations (9)

� game counters or sensors (2)


� trapping (3)

� collection and identification of scats (13) and collection of hair (4), and

� other techniques used included dusting with fluorescent pigment and pitfall traps.

4.3.7 Quantification of the Negative Impact of the Road and Traffic

The negative effect of the linear infrastructure or traffic on wildlife was evaluated in 50 per cent of the 59 studies. The majority of these referred to previous studies or used general ecological principles to predict that the linear infrastructure was likely to reduce connectivity.

4.3.1 Assessment of Factors Influencing Rate of Use

Most studies included an assessment of factors influencing rate of crossing (49 of the 59 studies). Of these 49 studies, 19 used a quantitative approach to assess the influence of different parameters on the rate of crossing (e.g. dimensions of structure, rate of use by humans, traffic volume, and presence of vegetative cover at the structure entrance), typically through a correlative or regression approach. The remaining studies that made conclusions about the factors influencing rate of use, typically including qualitative judgments that incorporated the results of other studies.

A range of variables was identified as influencing the rate of use of the mitigation structures by wildlife. Common themes appear consistently. Variables that appear to positively influence rates of use include:

� abundant and high-quality habitat near to the entrance of the structures

� dirt or ‘natural’ floors

� large ‘openness’ ratios (length x width x height of underpass)

� absence or low rate of use by humans, and

� presence of ‘furniture’ such as logs, rocks and vegetation on or in the structure.

However, it should be noted that the direction and magnitude of the effect of these and other variables are likely to be species or species-group specific, and were shown to vary from location to location. Furthermore, the effect of correcting for local abundance at each


crossing structure (sensu Clevenger et al. 2001) will further refine the identification of important variables.

4.3.1 Extrapolation or Study of Effect at Population and Community Levels

Only two studies clearly demonstrated that the use of the crossing structures led to a measurable increase in the viability of the population (Mansergh and Scotts 1989; van der Ree et al. in review). The second study (van der Ree et al. in review) used a 20-year data set that formed part of the annual monitoring of populations of Mountain Pygmy-possums and built a population model to evaluate the effects of the road without a tunnel, a road with a tunnel and a population distant from the road. The change in viability occurred because the tunnel allowed males and young animals to disperse from areas occupied by females prior to the onset of winter. This in turn lowered population density, reduced competition and allowed females to increase body weight before the onset of winter, increasing the rate of over-winter survival. Eight other studies made reference to population-level effects or implied in their discussions that the crossing structure probably increased viability and prevented a population sink.

An analysis of community or combined multi-species responses to mitigation systems has not been conducted. While numerous studies identified that multiple species used the structures, there was no evaluation at a community level with interactions among species. For example, what might the consequences be for a predator if the prey species are unable to cross the road?

4.4 Discussion

4.4.1 Aim and Effectiveness of Wildlife Crossing Structures

Evaluating the success of mitigation can be achieved at two different levels, with the first relating to delivery of suitable mitigation structures or systems and the second relates to the use of the structure by wildlife. The first measure of success, which we have termed compliance success, evaluates whether the final product has been constructed and delivered to the satisfaction of the expert biologist who may have proposed some of the original concepts. Compliance success was not reported on in any of the publications we reviewed. However, it


is critical because small details may render an otherwise suitable mitigation structure unusable. The multiple stages from original concept, to design, tendering, construction and delivery involve different people, each with different levels of skill and interest in increasing permeability for the target species. There is a great risk that communication failure between biologists at each stage between concept and construction will result in slight modifications at each stage, which cumulatively may result in vastly different structures than originally intended. Figure 15 highlights the potential for this when the structure intended for wildlife passage is inundated with water while the structures intended for drainage remain dry. There needs to be a process in place to firstly ensure this does not happen and then if it does occur, there is recourse to rectify the problem.

Figure 15. Culverts for drainage (3 pipes on left) and movement of Brushtailed

Phascogales (box cell culvert on right)

The second measure of success relates to the function of the mitigation strategy at reducing the effects of fragmentation of habitat. Forman etal. (2002b) proposed that the overall objective of wildlife crossing structures is to ‘increase the permeability of a road corridor’ (p. 161). They list a series of six criteria against which to measure effectiveness, namely:

1. reduce rates of road-kill

2. maintain habitat connectivity

3. maintain genetic interchange


4. ensure biological requirements are met

5. allow for dispersal and re-colonisation, and

6. maintain meta-population processes and ecosystem services.

We propose that a fundamental measure for the effectiveness of wildlife crossing structures is an increase in the viability of local populations or prevention of likely reduction in viability (in the case of a road widening or upgrade). Implicit in the analysis for each species is the assumption that multiple species also need to cross the linear infrastructure. However, we were unable to find a study that undertook a gap analysis to identify all possible species in an area that may need to use the structures but were not detected.

According to at least some of the six criteria proposed by Forman et al. (2002b), all of the studies we reviewed were successful at the level of the individual animal. The majority of wildlife crossing structures monitored increased the permeability of the road by allowing individual animals to move safely across the road. In this sense, wildlife crossing structures are generally successful at reducing the fragmentation effects of roads and other linear infrastructure for the individual, at the time it was recorded using the structure. However, it has been noted that use of structures does not necessarily equate to conservation gain (Ng et al. 2004). In other words, a key question is whether the mitigation efforts are sufficient to decrease the potential risk of population extinction to an acceptable level? This needs to be considered in the context of the target species being investigated. There are also relative risk questions between different species.

From a conservation point of view, the risk of not evaluating effectiveness at the population level is that we may continue to install mitigation measures that only partially decrease the barrier effect without a substantial-enough increase in population viability. From the perspective of road builders, it is important that the mitigation measures increase population viability as effectively as possible, especially given the significant amounts of money spent to install some types of mitigation measures. Road funders, road owners and road construction and management agencies are also under increasing pressure and legislative requirements to ensure that they do not further endanger species, especially those facing extinction.


4.4.1 Study Design

Most studies that evaluated the effectiveness of wildlife crossing structures lacked sufficient replication to quantitatively evaluate the influence of factors on the rate of use. Even Yanes et al. (1995) who investigated the rate of wildlife crossing within 17 tunnels in Spain lamented that their small sample size prevented them from drawing conclusions about the importance of particular design features for wildlife. Adequate replication is critical because the natural environment is so variable and this influences the rate at which structures are used. An unrepresentative picture may be obtained if all the structures are coincidentally placed in areas with high or low population sizes, with high or low densities of predators, with high or low density of geographic or landscape features that encourage or discourage use. Greater levels of replication are required to accommodate these systematic and random variables. Designing studies with greater inferential strength are not cheap, but the reliability and usefulness of a rigorous, well-designed study will reduce the need to repeat the study (Roedenbeck et al. 2007)

The mitigation works and evaluation must have a clearly defined and measurable goal. The six criteria proposed by Forman et al (2002) and our goal ‘to increase population viability’ should be seen as guiding principles only, and not the actual goal against which success can be measured. The goal for each project must be specific to the location, species of concern and nature of the problem. We recommend use of the ‘SMART’ approach (Specific, Measurable, Achievable, Realistic, and Timeframed) to set a specific goal and thus facilitate more effective evaluation of mitigation measures.

An ecological goal for a road through habitat listed under the EPBC Act might be to ‘maintain the risk of extinction to less than 5 per cent over the next 100 years’. Alternatively the goal could be to maintain connectivity for a species or community by ensuring that ‘more than 90 per cent of individuals that attempt to cross the road do so successfully’. Examples of other goals may include ‘a 90 per cent reduction in road-mortality within two years after mitigation’ or ‘within five years of construction, 75 per cent of all individuals that enter a culvert will pass through’. The identification of specific goals for each project is likely to alter the emphasis of the mitigation. In one area the focus may be on reducing road-kill, while for another species it may be on maintaining daily movements.


As a precursor to setting realistic and sensible goals there is a need to more fully identify and quantify the negative effect of roads and traffic on flora, fauna and ecological processes. The number of animals killed after collision with vehicles is clearly a major issue and a cause of concern for both human safety and conservation. Nevertheless, road-kill is just one aspect of the negative effect of roads and there are likely to be many other direct and indirect effects of roads and traffic that need to be considered. Therefore, studies evaluating the effectiveness of mitigation measures should not rely solely on measuring the rate of road-kill as an index of crossing-structure success. For example road-avoidance by male mountain pygmy-possums was used by Mansergh and Scotts (1989).

Ironically, the absence of species in the road-kill tally may not mean that the road is not affecting populations or that the mitigation measure is successful. It may mean, for example:

� the vegetation in the area is not a suitable habitat

� other environmental effects of the road are occurring such as traffic noise, light or chemical pollution are deterring animals from even reaching the crossing structure

� small animals may be unidentifiable after collision

� a proportion of some species may leave the road surface and die in adjacent vegetation, or

� in the worst case, the local population may have become extinct or sufficiently rare that it can no longer be detected.

This highlights the need for good scientific data on aspects other than just the rate of detection within a structure. For example, the reported 64 per cent reduction in the road-kill on Highway 441 in Florida USA (Dodd et al. 2004) may be due to successful mitigation, but there could also be other factors at play. Information on the status of populations in adjacent habitats would help to elucidate causes in reduction in mortality and confirm that it is due to fewer animals accessing the road.

The evaluation of wildlife tunnels or overpasses should consider how post-mitigation vital rates (e.g. dispersal, gene flow, survival) compared with both the pre-mitigation situation and the non-road situation and evaluated using a population viability model. If the age or sex structure, survival, patterns of dispersal, and gene flow before mitigation are not


known, then it is impossible to assess whether these parameters have improved after mitigation. It would be insightful to understand the pre-development population dynamics. For example, data on the effectiveness of any existing wildlife movements or crossing structures in the vicinity of the proposed major linear infrastructure would be most useful. We are not suggesting that a detailed field assessment of the actual impacts of every road be undertaken prior to mitigation, however some useful information may already exist in some situations.

The cost to retrofit mitigation structures to existing roads to make them more permeable is generally prohibitive. Therefore it is essential that appropriate mitigation measures be incorporated into the construction of new and upgraded roads. However, some existing structures (e.g. drainage culverts) may be relatively easily and cheaply modified to make them more fauna-friendly. This may include modifying entry and exit areas to be more natural, the installation of timber beams along the walls and other ‘furniture’ to provide suitable substrates for movement and protection from predators. Irrespective of the modification, it is essential to collect the required baseline data, to thoroughly document and understand the ‘before’ situation, prior to the commencement of capital works.

A poorly designed or conducted assessment of the effectiveness of wildlife mitigation structures could be inefficient and counter-productive (Roedenbeck et al. 2007). Poor assessment could result in installing insufficient measures, or of the wrong type, or in the wrong places (or, conversely, it could result in installing too many, wasting considerable amounts of money). It is still unknown how many structures are required to reconnect two populations of a species with small home ranges. Is one tunnel per km sufficient, or one per 500 metres or one per 100 metres? The answers to such questions relates back to the identifying the nature and extent of the likely problem. If the connections are for gene flow, then one effective dispersal per generation may be sufficient. If the connectivity is to allow the annual migration of a certain species, then multiple structures at frequent intervals may be required.

The mean duration of Australian studies appeared to be considerably shorter than those from overseas. This was 4.2 months and 2 years, for Australian studies and overseas respectively (excluding the 20-year Australian study of Mountain Pygmy-possums by van der Ree et al. in review). Numerous studies have shown an increase in the rate of use of


mitigation structures over time as animals become accustomed to the structures, as disturbance due to construction is rehabilitated and as vegetation cover increases. Furthermore, seasonal variation in rate of use is also evident, and this is unlikely to be detected in short studies. There appears to be a trend for longer-term studies being undertaken overseas (over a number of years). These longer studies could have the following components of pre-mitigation studies, to develop a baseline; monitoring during construction; and then post construction monitoring.

Remotely-triggered cameras and tracking pads were the most commonly used techniques to survey use of crossing structures, and while they are effective at detecting large species and those with diagnostic tracks, they are less efficient at detecting smaller and more cryptic species. These species with lower rates of detection are also likely to be affected by habitat fragmentation, and thus greater effort needs to be expended to evaluate the suitability of these structures for them. Possible survey techniques include pitfall trapping within mitigation structures, genetic techniques to detect dispersal across structures, and tracking pads that have a finer substrate (while protected from larger animals) to record footprints, such as ink or marble dust.

4.4.1 Improving Knowledge Transfer

There can be considerable cost to construct some forms of mitigation (e.g. up to $US16 million in 2002 to construct a tunnel (Forman et al. 2002b). Thus there is an imperative that effectiveness be analysed appropriately and fully evaluated. In this review we assessed information published in the refereed scientific literature and in reports from Australian consultants. Furthermore, the primary literature sources that are most accessible to road engineers and consultants around the world are likely to be international peer-reviewed journals and some recently completed agency reports. For example, three of the recently published studies that we reviewed (Cain et al. 2003; Taylor and Goldingay 2003; Ng et al. 2004) all cited three of the earliest papers published in journals on the use of underpasses (Reed et al. 1975; Reed 1981; Singer and Doherty 1985). Therefore, we believe that studies on the effectiveness of wildlife crossing structures should be submitted to peer review in appropriate scientific journals. We believe that the additional costs involved in writing up the findings for publication in a reputable journal should be factored into commissioned studies. An


alternative suggestion is that agencies wanting to build the linear infrastructure should stipulate that the findings of commissioned mitigation reports should also be prepared for submission to scientific journals.

To further improve the efficient and accurate transfer of knowledge, we suggest a series of minimum criteria be reported in all studies. This is critical for the reader to gain an understanding of:

� overall configuration of the linear infrastructure

� the surrounding vegetation

� road conditions, and

� how the mitigation structures themselves might influence target species and habitat.

The overall configuration of the linear infrastructure includes the number and width of vehicle lanes or powerline clearings, particulars on service and access lanes, and details of central median. The presence and type of vegetation adjacent to the road (and within the central median) may act as potential habitat and therefore should be described. The road traffic conditions require clarification. Pertinent factors include mean vehicle speed, and variations in vehicle speed, traffic volume, and times of peak traffic flow.

The physical characteristics of each structure should be spelt out clearly and include dimensions (length, width, height); cross-sectional shape (e.g. round, rectangular); intended function (drainage, wildlife passage); and mode of construction and materials (e.g. pre fabricated concrete box culvert). This is important to avoid potential confusion due to inconsistent nomenclature across regions and studies. We also suggest adopting standard definitions of structures as outlined in Table 1. We would suggest that the width of the road and clearing and structure length be detailed separately.

4.4.1 Importance of Study of Animals in Adjacent Habitat

The rate of use of crossing structures is likely to be partly related to the abundance of animals in adjacent habitat (e.g. Yanes et al. 1995). Thus, studies that draw conclusions about the suitability of certain types of structures without considering the availability of animals to actually use the structure may give an incomplete assessment of suitability.


One interesting approach calculated an expected rate of crossing based on animal abundance in adjacent habitat and compared the expected rate of use to that observed (Clevenger and Waltho 2000; Clevenger etal. 2001). At the very least museum and wildlife atlas records can provide a list of species that probably occur in the study area. However, this should be considered the very minimum at which the pool of potential species be estimated. What is required to more fully elucidate the effectiveness of certain structure types are crossing rates relative to local abundance, with local abundance preferably estimated at the time of monitoring.

Similarly it is critical to understand the way in which a tunnel or overpass is being used. In particular, the consequences for population viability may differ if the structure is used for occasional or dispersal passage or daily as movement within a home range. For example, a concurrent radiotracking study of tunnel crossings under Highway I-75 in Florida USA for the Florida Panther found that of the 10 reported crossing only two individuals were involved, and use was related to the home ranges of the panthers (Foster and Humphrey 1995). For some migratory species, the direction of travel and time of year is strong evidence that the crossing structure is used in migration. Similarly, the daily use of underpasses by mountain goats to access a salt lick (Singer 1978; Singer and Doherty 1985; Pedevillano and Wright 1987) is convincing evidence how the structure is being used.

Some studies lacked strong evidence, and it was unclear how the structure was being used and the number of individuals using it. For example, the number of tracks of a certain species does not necessarily equate to the total number of individuals using the structure and thus studies should be always viewed with initial circumspection. It may be that a dominant individual has established a territory and frequent use of the structure prevents access by other individuals. Therefore in such circumstances mitigation measures to facilitate crossing should be placed at a frequency that corresponds to the spatial scale over which the target species moves (McDonald and St Clair 2004). Finally, recording the presence of an individual within a structure (e.g. recording footprints at a tunnel entrance) does not always equal a successful crossing. Therefore, the minimum standard for recording a successful crossing might be a set of tracks travelling in the same direction recorded at both entrances to the tunnel or overpass (e.g. Gloyne and Clevenger 2001; Ng et al. 2004).


4.4.1 Conclusions

Many road construction agencies around the world are constructing and modifying roads to have less environmental impact. Major linear infrastructure capital works consume huge amounts of money. Notwithstanding this, the amount spent on mitigation is relatively small compared with the overall construction and maintenance budgets of state and national road agencies. However the amount spent on linear infrastructure is significant to when it is compared with how much is spent on the full range of general conservation actions. Thus it is imperative that mitigation dollars are spent wisely to achieve maximum benefit.

Because of the recent surge in research and expenditure on minimising the ecological effects of roads and traffic it is pertinent and timely to evaluate the effectiveness of mitigation measures and advise on the direction that future research and monitoring should take. The studies we reviewed clearly demonstrate that most measures designed to increase the permeability of roads for wildlife are successful at the level of the individual animal. The detection of an animal in a tunnel or overpass indicates that on the occasion it was detected it probably made it safely to the other side, and thus can be considered successful. However, the extent to which the population and local community has benefited from that successful crossing is unclear. There is insufficient information and analysis in the majority of studies to evaluate whether the viability of the population has increased to an acceptable level. Future research and monitoring should quantify the extent to which the structure (and series of structures) has enhanced population viability. It is possible, particularly for sedentary species, that the positive effect of a crossing structure is very localised and extends only a short distance from the structure. Without this knowledge, determining the number and spatial distribution of mitigation measures required to increase viability is difficult. Finally, the barrier effect of roads is just one potential impact on fauna, and mitigation that addresses this may only increase viability within the limits posed by other effects.


CO

ST

– B

EN

EF

IT A

NA

LY

SIS

5 Cost – Benefit Analysis of Habitat Fragmentation Mitigation Measures

Cost benefit analysis is a technique designed to determine the feasibility of a project or plan by quantifying its costs and benefits. Several factors need to be considered when attempting to provide a cost benefit analysis for habitat fragmentation mitigation measures so that the analysis is accurate, useful and replicable.

5.1 Paucity of Hard Data for Effective Financial Cost and Ecological Benefit

During this desktop study, it has become evident that there is very little hard data available to effectively assess the cost-benefit of habitat fragmentation mitigation measures. Of the 59 studies we reviewed, only two provided information on the cost of their mitigation measures. The only Australian example was Abson and Lawrence (2003b) for the bridge above Slaty Creek. This bridge was enlarged beyond what was required for drainage, and cost a total of $3 million.

There are two major components to this cost benefit analysis calculation. Firstly there is the cost of the linear infrastructure and mitigation structures. The second determinant is to measure the utility and benefit of the habitat fragmentation mitigation measures on a particular target species. Both components are difficult to calculate, having their own particular problems. We have ascertained that the information and opinions that are in the public arena are of variable quality, somewhat subjective, incomplete, somewhat unreliable, and potentially misleading. At this stage the information is certainly not of a quality or consistency to make it useful or useable for policy formulation. The reasons the data is so poor is because:

� there are comparatively few people involved in this work undertaking the analysis, thus the body of work is small

� much of the real, actual or final cost of the structures or the roads themselves is closely guarded business-in-confidence information, or that it takes years for the final cost of the roads to be totalled-up and become available, and

� the analysis is very complex, involving many variables.


5.2 Variables in the Cost – Benefit Analysis

Some of the critical variables include:

� the intrinsic value of the species being targeted (e.g. are we prepared to spend more money protecting the endangered Mountain Pygmy-possum versus the common and widespread species of rat or kangaroo?)

� habitat type

� character and topography through which the linear infrastructure passes

� number of mitigation structures installed along a given unit stretch of linear infrastructure (road, rail etc)

� size, complexity, functionality and aesthetics and other requirements demanded for the structures

� time road/mitigation measure were constructed (There also appears to be an increasing cost of structures – above and beyond the background increase in construction costs)

� lack of a standardised and meaningful measure of effectiveness, both in terms of use, as well as increases in population viability (which was evident in much of the literature reviewed)

� intensity of road use (e.g motorway to minor country road, the speed, and volume of traffic)

� characteristics of road and associated features (e.g. number of carriageways, width of carriageways, nature and width of verges and medians, road drains, road barriers and other road furniture)

� difference of requirements and cost between types of linear infrastructure (e.g. road vs railways)

� difficulty in measuring the ‘benefit’ to assemblages, communities and ecosystems and of the negative impacts and compounding feedback of not installing a given mitigation measure

� whether it is an upgrade of an existing road or a new road;

� regional difference in the cost and difficulty of conducting the construction, maintenance and upkeep i.e. urban, peri-urban, rural, remote, different States/Territories, etc

� construction techniques, economies scales etc, and


� design variability, and relative understanding, experience and expertise of client, designers, and constructors.

5.3 Accounting for the Differences in Cost

With the above variables in mind, it can be seen that performing a useful cost benefit analysis is very difficult. An example of the variability in the cost of mitigation structures is illustrated below (Table 5). The costing for a simple 3x3 metre box culvert on the Pacific Highway project is compared with a road project in the ACT. The first example is derived from RTA estimates (O'Donnell 2003) and the second costing was derived from SMEC internal costing estimates for a typical box structure on the Gungahlin Drive Extension Project in the ACT.

Care has to be exercised when considering costing for apparently similar structures. It is imperative to ensure that the basis of comparison is known and that all the input costs are known to achieve comparability. For example it is important to know fundamental factors including:

� length of underpass;

� locational issues, such as geotechnical or grounds factors/difficulties

� construction matters, such whether the installation is a relatively simple mass-produced prefabricated structure, or a specifically designed more complex structure, and

� what other works are included or excluded in the cost structure.

Also, variance in cost can be expected as structures become larger, more complex or when not of standard design.

Similarly, over the last few years construction costs have increased significantly. Industry margins on work often vary significantly, even between apparently similar jobs. Also, it must be realised that most quoted costs are estimates of cost. When final costs are ultimately determined the cost of individual structures can vary considerably, depending on the actual conditions or problems that were experienced on particular jobs, or at specific sites within the same project.

The example in Table 5 shows a difference in cost of $110 000 for a ‘standard’ 3x3 metre box culvert between two construction sites and over a 3 year period (2003 and 2006).


Table 5: Approximate Cost of Various Structures

Mitigation Structure Pacific Hwy 2003

(estimated average)1

Gungahlin Drive Extension 2005

(estimated average)2

Pipe Culvert underpass (1.8m diameter)

$50 000 5.3.1

Box Culvert underpass (3x3)

$225 000 $334 900

Arch Underpass (4.5 x 12)

$300 000 $1.7 million

Arch Underpass (8 x 21) $901 000 5.3.1

Arch Overpass, Twin

(8.5 x 15)

$1 000 000 5.3.1

Arch Overpass, Twin (8.5 x 15)

$1 200 000 5.3.1

1 O'Donnell J. (2003) Practical and Cost Effective Terrestrial Fauna Friendly Design on the Pacific Highway. Powerpoint Presentation. 2 Internal SMEC costing estimates for a box culvert and arch underpass on the Gungahlin Drive Extension Project in the ACT.

We estimate that the current unit cost of a glider pole (without guardrail) would typically be between $3000-$5000, depending upon a number of factors such as location and difficulty of erection.

There are also significant additional costs in retro-fitting mitigation measures after a road construction period has passed. For example, the estimated cost of installing each rope ladder on the Pacific Highway project was approximately $23 000 during the construction phase, whereas an estimated cost of installing a rope ladder of approximately 100 m in length on the Hume Freeway in north eastern Victoria after construction in 2007 was $70 000 – $90 000. Nevertheless, the cost to retrofit existing mitigation structures with features that enhance their rate of use by wildlife, such as log shelving above water levels or revegetating the entrances or exits to culverts, is relatively small and may provide significant cost-benefits.

The approximate cost of highway construction per kilometre of various types of road in 2006, from the Rawlinson Australian Construction HandBook3, is given below in Table 6. It should be noted that these figures are highly dependent on a number of factors and are only indicative.


Table 6: Approximate Highway Construction Costs in 2006

Road Type NSW QLD

Country Highway with Shoulders (Two Lanes)

$480 000/Km - $530 000/Km

$255 000/km - $305 000/Km

Country Highway with Shoulders (Three Lanes)

$715 000/Km - $785 000/Km

$365 000/Km - $445 000/Km

City Highway/ Freeway with Median Strip and Emergency Lanes (Duplicate Two Lanes)

$2 300,000/Km- $2 330,000/Km

$1 660,000/Km - $1 960 000/Km

City Highway/ Freeway with Median Strip and Emergency Lanes (Duplicate Three Lanes)

$2 480 000/Km-

$2 780 000/ Km

$1 810 000/Km - $2 110 000/ Km

Assumption: Prices include minimal cut and fill but exclude lighting and drainage.

Estimates may be 40 per cent either side of these prices.

3 Rawlinsons (2006), Rawlinson Australian Construction HandBook, Rawlhouse

Publishing Pty Ltd: Perth, Western Australia.

5.3.1 Cost – Benefit Models

For this form of analysis to be potentially useful in practice it requires the development of an agreed functional model, including what are the significant variables. The model would have to attribute ‘equivalent economic value’ to the various ecological functions, and apply statistical standardisation of the variables. It would be necessary for the model to be clear and transparent about on the assumptions and caveats and the confidence levels. Development of such complex multi-variable models would have to done in close collaboration with stakeholders to ensure it had all of key variables, was linked appropriately, with agreed caveats and weightings. Then the model would have be tested against reality and refined as deemed appropriate.


6 General Principles

After conducting this review we believe that there are some core principals applying to the mitigation of impact of the fragmentation caused by major linear infrastructure. We believe that if these principles are applied it will greatly reduce the impact. We suggest that the Department of Environment, Water, Heritage and the Arts send the principles to parties involved in the funding, commissioning, reviewing, specifying, designing and constructing of major linear infrastructure commending them for application for new linear infrastructure and the expansion of existing linear infrastructure. These principles are:

� Fragmentation is only one of the effects of linear infrastructure.

� Avoid environmentally sensitive areas.

� Identify the nature of the issues.

� Better to connect than fragment.

� Identify the goals for mitigation (use the SMART technique).

� Design mitigation structures for faunal groups, communities and ecosystem processes.

� Mitigation structures should be for a wide range of species.

� Understand conditions and populations adjacent to structures.

� Use and support targeted research.

� Monitoring should be an integral part of the construction and management process.

6.1 The Principles

1) Fragmentation is only one of the effects of linear infrastructure

The deleterious effects of linear infrastructure on the environment are numerous and varied. The fragmentation of habitat and the creation of barriers caused by linear infrastructure may reduce landscape connectivity, and in turn, reduce the population size of species of biota. The majority of mitigation efforts have focussed on increasing the


permeability of roads and reducing rates of mortality. Other potential effects of roads may also have significant effects on the viability of local populations, and should also be considered.

2) Avoid environmentally sensitive areas

The principle of avoiding environmental damage at the planning stages is a high priority. By avoiding environmentally sensitive areas many of the mitigation and compensation measures may not be necessary. In cases where avoidance is not possible, mitigation options should be considered.

3) Identify the nature of the issues

Does linear infrastructure pose a threat through the fragmentation, loss or degradation of habitat, increased mortality, and risk to human safety and financial costs? Are there other issues that have been identified? It is critically important that the issue be clearly articulated so appropriate mitigation measures can be identified. For example, a solution to minimise rates of mortality may be to construct exclusion fencing, however this could also increase the fragmentation effect.

4) Better to connect than fragment

Even if the effectiveness of a certain mitigation structure is unclear, or the full extent of a barrier has not been quantified, it is better to attempt to mitigate than not mitigate at all. It is more cost-effective to install mitigation structures during the construction of larger projects than attempting to retrofit structures. The precautionary principle suggests that it is better to restore and maintain connectivity than not.

5) Identify the goals for mitigation

To fully evaluate the success of any mitigation measure the goal of the mitigation needs to be clearly and carefully defined. Using the SMART technique the objectives need to be:

� Specific

� Measurable

� Achievable

� Realistic, and

� Time-framed.

Adopting this approach to monitoring will enable a more targeted and specific evaluation of the effectiveness of mitigation. In the absence of


this approach, it is likely that the decision to modify mitigation structures or to cease monitoring will be made on an ad-hoc and subjective basis. Clearly defined criteria and measures of success will provide more certainty to road construction and management agencies that have to address various legislative or other requirements. It will also assist them to design and deliver mitigation works that substantially contribute to the conservation of biodiversity.

6) Design mitigation structures for faunal groups, communities and ecosystem processes

It is important to recognise that multiple species are likely to be affected by the linear infrastructure. Individuals and species do not operate in isolation but interact with each other to varying degrees. Furthermore, it is likely to be more cost-effective to build individual multi-use structures that serve multiple species, rather than build more structures narrowly designed for only one or two species.

7) Mitigation structures should be for a wide range of species

Another desirable mitigation option is to facilitate the movement of other creatures that are not commonly catered for (normally the focus is on mobile terrestrial mammals) eg. flightless beetles, slugs and underground fungi also need to move across the landscape.

Within Australia, approximately 113 faunal species or groups have been detected using underpasses and overpasses, approximately half comprising mammals, the remainder including amphibians, reptiles, birds and a very small number of invertebrates. The focus to date has been on mammals because they are relatively easy to study, and are more ‘charismatic’ than other groups. The effectiveness of mitigation structures for other groups, such as reptiles, amphibians and less mobile birds and invertebrates needs to be investigated, as these groups are also likely to be significantly impacted by the fragmentation of habitat.

8) Understand conditions and populations adjacent to structures

It is important to determine the number of species in an area adjacent to a crossing to gain an insight into which species are ‘available’ to use that structure. The rate of use by a species will be related to its abundance and density in the area. It should be noted that otherwise well designed structures might be sub-optimal if there are low


population densities in the vicinity. Therefore, to be able to compare the effectiveness of a structure it is important to measure the local species abundance, and then assess the relative crossing rate. Surveys of local abundance (not just occurrence) need to be undertaken within the vicinity of each structure to estimate population density or calculate an index of population density, simultaneously to the assessment of structure use.

9) Use and support targeted research

The research and monitoring of the use of mitigation structures conducted to date has typically focused on documenting the presence of species and relating presence or abundance to features of the landscape, habitat or mitigation structure. Crossing structures of major linear infrastructure do work, the questions are how well they work and how to improve effectiveness. We recommend that future research should additionally focus on answering higher-order questions, such as:

� What rate of crossing is required to maintain the selected target populations?

� What is the rate of crossing across linear infrastructure in the absence of mitigation structures? Under what conditions (e.g. road width, traffic volumes, population density) are crossing structures required?

� Did the animal detected make a full crossing of the structure?

� What is the precise purpose of the structure (e.g. dispersal, daily foraging etc)?

� What proportion of dispersals across the infrastructure is successful?

� Has the mitigation improved population viability to a satisfactory level?

� What other species (e.g. invertebrates, reptiles, amphibians) or ecosystem services (e.g. dispersal of seed, nutrient cycling etc) require mitigation?

� Other research questions are detailed further in Section 7.11

10) Monitoring should be an integral part of the construction and management process

Finally, monitoring needs to be more than simply the last task that needs to be undertaken before the project is completed. Research and


monitoring should be integral to the project from the outset. It should be a formal adaptive experimental management approach, where results are fed back into the design and management of existing structures and used as a basis to inform the design of future mitigation systems. While this feedback happens now, it is ad-hoc and the process can be made more objective, transparent and effective. The road-project by road-project approach to monitoring and evaluation limits the overall inferential power of studies. It may be more cost-effective to study a larger number of structures simultaneously, using standard and consistent methods, than a series of multiple studies conducted by different people using different methods.


CO

NC

LU

SIO

NS

AN

D R

EC

CO

ME

ND

AT

ION

S

7 Conclusions and Recommendations

7.1 Effects from major linear infrastructure

The following are a summary of the many direct and indirect ecological effects of major linear infrastructure:

� loss and degradation of habitat

� incursion of weeds, disease and feral animals

� direct mortality of wildlife, due to collision with vehicles

� disruption of movements due to the creation of barriers

� altered microclimatic conditions, and

� changes to the acoustic and light environment.

7.2 The need for crossing structures

Many landscapes across Australia, and particularly those where major linear infrastructure are planned, are already highly cleared and fragmented. In these situations populations of biota are at an increased risk of extinction due to smaller population sizes and restriction to certain parts of the landscape. While there is a general concept or rule that one effective migrant per generation may be sufficient to prevent inbreeding effects, a greater rate of dispersal among sub-populations is required to re-populate or prevent a local extinction. Additional fragmentation, in the form of new or wider roads, is likely to further exacerbate these pressures. We recommend that all linear infrastructure projects avoid fragmenting the landscape where possible, thus avoiding any need for mitigation works. Where additional fragmentation is unavoidable, road planners, designers and biologists meet early in the process to identify, design and implement the best mitigation strategy. This is important not just for rare or endangered species, but also for species that are now relatively common. It is possible that road mortality and / or barrier effects may reduce the likelihood that even moderately abundant species persist into the future (e.g. Ramp and Ben-Ami 2006).

7.3 The use of wildlife crossing structures

The majority of wildlife crossing structures installed in Australia and overseas do increase the permeability of the landscape for a wide variety of species. They are successful at facilitating the crossing of


linear infrastructure by wildlife. Numerous individuals of many species were detected moving safely from one side of the road to the other.

7.4 The efficacy of wildlife crossing structures

The effectiveness of wildlife crossing structures and other mitigation systems is more difficult to answer because specific goals for the mitigation have not been clearly articulated a priority. A significant question that remains unanswered is: ‘How effective are the wildlife crossing structures at increasing population viability?’ The ultimate test of mitigation success will be whether the population effected by the road still occurs at the same or higher density in the vicinity of the road in 10, 50 or 100 years time. Similarly, the ability of the mitigation to maintain higher-order levels of complexity, such as community structures and ecosystem processes is unclear. However, structures that facilitate crossing by wildlife are likely to contribute positively towards achieving these goals. In this context, the implication of not evaluating effectiveness appropriately is that mitigation efforts may be insufficient.

7.5 Pre-Construction Studies

It is important that preliminary ecological studies of the potential impact of possible new major linear infrastructure be conducted at the earliest possible moment. This is to give the best opportunity to design and conduct an adequate and useful base line study. Doing so will greatly enhance all subsequent ecological and mitigation work.

To achieve this early involvement requires arrangements to ensure that environmental and ecological staff are formally alerted and consulted at the concept stage of every major linear infrastructure project.

These preliminary ecological studies should be properly designed, and adequately funded.

7.6 Monitoring and evaluation

The first step in evaluating a structure is to determine compliance success. Does the structure or system that was delivered meet the expectations of the original concept and fulfil the design criteria. In other words, does the mitigation look as it is supposed to look? Biologists with expertise in the habits of the target species should be engaged earlier in the process and consulted at each phase of project


delivery to ensure that the designs are not modified beyond which the target species is no longer likely to use it. Ongoing maintenance of the structure or system must also be planned and budgeted for and include such things as inspection of fencing, removal or replacement of vegetation as required, replacement of furniture etc.

The number of species using structures and rate of their detection is a first-step in evaluating the use and effectiveness of wildlife crossing structures. There are a range of approaches to studying road effects and success of mitigation that are either manipulative or non-manipulative (Roedenbeck et al. 2007). These range from before-after-control-impact (BACI) designs (with greatest strength of inference) to less robust control-impact or just impact designs. However we would council that the most effective survey and monitoring techniques are not necessarily the cheapest. The type and level of inference that can be gained from a well-replicated BACI design is far superior to a cheaper study of what animals were found in structures after construction (Roedenbeck et al. 2007). Typically the benefits from more expensive designs are that managers will be able to accurately determine if their mitigation measures are having the desired effect according to the specific pre-determined goals. Undertaking research and experiments that are more reliable will require more planning than currently available under existing road design and construct models. In some situations, the necessary ‘before’ studies may take a number of years to complete. For this to occur the issues need to be identified and research planned with sufficient lead time prior to construction, and probably earlier if research outcomes are to inform road design.

A hierarchy of monitoring and evaluation range from a simple assessment of compliance success and which species use the structure to more complicated questions about effectiveness, gene flow, population viability. In certain limited situations it may be possible to expend less effort on evaluating use in situations where previous studies have already demonstrated that species x will use a similar type of mitigation structure. However, at the current stage of the science, we still believe that a more rigorous assessment of effectiveness, with studies of animal abundance in adjacent habitats is required to improve our understanding of current best practice. Funding and approval agencies must recognise and support the move towards more in-depth and longer-term studies that more rigorously answer the questions that road designers and builders regularly ask of biologists.


A key step in evaluating mitigation effectiveness is to assess the outcomes against a clearly defined goal. We recommend the use of the SMART approach, where each goal is specific, measurable, achievable, realistic and timeframed. This approach allows project and location-specific goals to be identified and evaluated.

We recommend that monitoring be an integral part of the construction and management process for all major infrastructure projects, preferably by being formally required in the specification of works.

7.7 Information dissemination

Road ecology is an emerging science and numerous groups around the world are undertaking research to further quantify effects of linear infrastructure and evaluate mitigation. Australian science is certainly at the forefront in this international field with respect to understanding the effects of roads, but we appear to lag behind when it comes to evaluating effectiveness of mitigation systems (e.g. see publications in Irwin et al. 2003; Irwin et al. 2005). This discrepancy is driven in part by the North American and European focus on reducing rates of collision rates between large animals and vehicles (i.e. human safety). A greater focus on publishing results of evaluations in peer-reviewed journals will, over time, improve scientific rigour. However, as much of the current monitoring is conducted by ecological consultants, there is little time or resources available to prepare manuscripts for submission to a peer-reviewed journal. Funding agencies should consider including a commitment to publish when awarding contracts and provide final payment upon submission or final publication. Not all monitoring will or should be published in journals, and a central clearing house where all ‘road ecology’ papers get submitted and made freely available online should be considered.

7.8 Cost and benefit

As mentioned in the report, major linear infrastructure capital works consume huge amounts of money and, comparatively, the money spent on mitigation is relatively small, however it is not insubstantial. Thus it is imperative that mitigation dollars are spent wisely to achieve maximum benefit.

Quantifying the cost and the benefit of mitigation measures is not easy. Currently there has been minimal work done in this area. At this stage


the quality and consistency of data is insufficient to properly inform policy formulation. Therefore it may be beneficial for an appropriate agency to support a comprehensive cost benefit analysis of a habitat fragmentation mitigation measures.

Given the paucity of definitive work in this area and absence of an agreed model it is recommended that in the first instance the Department convene a workshop of key stakeholders and relevant experts to develop an agreed, standardised methodology for conducting a productive cost benefit analysis of habitat fragmentation mitigation measures. Following this, the Department would be in a much better position to develop the economic/benefit model and undertake appropriate studies

7.9 General principles and national guidelines

There appears to be considerable good intent and willingness by parties involved in the construction of major linear infrastructure, including the commissioners and designers of such works, to properly address habitat fragmentation issues. However, the project-by-project approach, and the involvement of a large group of stakeholders (from concept to construction phases) results in mitigation works of variable amounts and quality.

We believe that a set of national guidelines would assist in achieving consistently higher standards across Australia, and could, over time, develop from the general towards more specific guidance on the type and placement of mitigation systems. The promulgation of agreed national guidelines would be very helpful to all.

We recommend that the General Principles enunciated earlier in this report be used by the Department as a basic starting point for national guidelines. We suggest as a start that the general principles be sent to all parties involved in the funding, commissioning, reviewing, specifying, designing and constructing of major linear infrastructure.

In conjunction with the initiative on cost and benefit (mentioned at 7.8) the Department of Environment, Water, Heritage and the Arts may wish to also concurrently convene a workshop. A workshop that includes the widest range of interested parties, charged with moving towards refining and enhancing the principles into a more comprehensive consensus national guidelines.


7.10 Standard definition of terms

As mentioned earlier in the report we believe it would be useful if there was standardisation of terminology. We commend the schema adopted in this report. Developing the guidelines outlined above will also assist considerably towards achieving uniformity.

7.11 Future research

As noted in this report there is considerable paucity of hard data on which to develop soundly grounded policy and practice. Thus there is a need for appropriate research. We believe this research should be targeted to specific issues and needs. To be effective and useful, the research needs to be of the highest quality and properly designed, and adequately funded.

Research into the effects of major linear infrastructure is currently funded in an ad-hoc way in Australia. It would be desirable for a more formal arrangement to assist in the funding of targeted research. Programs to commission targeted research occur in other areas, such rural research and health research, through special targeted research programs run by the Department of Primary Industry and the National Health and Medical Research Council, respectively.

Potentially this special targeted research program in linear infrastructure could be funded by the funders and commissioners of linear infrastructure to address their key unanswered questions. Of course there would have be formal arrangements to decide who would fund, coordinate and commission such a research program.

However, to start the ball rolling, it is recommend that the Department of Environment, Water, Heritage and the Arts convene an exploratory high level meeting or workshop with the key stakeholder to discuss this concept.

Some future targeted research could include:

� What are the minimum rates of crossing required to prevent local extinction, and how frequently do crossing structures need to be placed?

� What is the precise purpose of specific structures (e.g. dispersal, daily foraging etc) and to what extent do they meet that purpose?

� What other ‘at-risk’ species not using crossing structures should we be concerned about? (e.g. invertebrates, reptiles, amphibians?).


� What are the consequences of current levels of road-mortality? What data sets on rates of road-kill exist and what do they show? Are targeted road-kill surveys required to elucidate population-level effects?

� Where are successful and unsuccessful crossings occurring? What factors influence rates of crossing and mortality?

� What proportions of dispersals across the infrastructure are successful?

� What is the extent of the barrier effect for species currently observed within crossing structures? In other words, are we currently catering for species experiencing the greatest levels of fragmentation?

� How do we determine what is a satisfactory risk of extinction for a target species?

� Will (has) the mitigation structure improved population viability to a satisfactory level?

� What other species less often considered (e.g. invertebrates, reptiles, amphibians) or ecosystem services (e.g. dispersal of seed, nutrient cycling etc) require crossing structures or other forms of mitigation?

� Explore the establishment of a research hub that includes government, biological and ecological research agencies/universities, road construction groups, planners, traffic safety analysts etc. Similar groups have been formed in the USA (e.g. Road Ecology Center at University of California: www.roadecologycenter.org; Western Transportation Institute at Montana State University: www.coe.montana.edu.wti; Center for Transportation and the Environment at North Carolina State University: www.cte.ncsu.edu) and they each provide a focus for slightly different research directions.

� How effective are mitigation strategies across non-road linear infrastructure?

� Cost Benefit Analysis.


SU

MM

AR

IES

OF

RE

LE

VA

NT

LIT

ER

AT

UR

E

8 Summaries of Relevant Literature

Papers and articles on mitigation structures are basically of two types. The first type we have classified as typically being refereed journal articles or consultants reports that detailed the results of research investigating the use and effectiveness of different mitigation options. The second were ‘best practise’ manuals and reports that provided varying levels of detail on the appropriate mitigation options for a certain species, location or context.

The criteria we used to consider these documents were:

1. Publications needed to present data on use of mitigation structures by wildlife. Publications that discussed potential structures, or vaguely summarised other reports were not included.

2. Opinions and assertions that were not supported by any data were not included.

3. Primary literature sources were reviewed (where possible) (i.e. summaries and other reviews to be avoided because of the tendency to misinterpret second-hand information).

4. All Australian studies were given priority.

5. For international literature only those published in refereed scientific journals were given priority consideration.

6. International grey literature (reports, conference proceedings etc.) were included if they represented, in our opinion, an advancement in approach or significance.

8.1 Australian Refereed Journal and Book Articles

Taylor, B.D. and Goldingay,R.L. (2003) Cutting the carnage: wildlife usage of road culverts in north-eastern New South Wales. Wildlife Research 30: 529-537.

Location of study: Brunswick Heads Bypass on Pacific Highway, NE NSW, Australia

Habitat: Assessed: dominated by swamp sclerophyll open forest and woodland. West side of bypass highly disturbed, extending for 5 metres beyond fence and dominated by exotic palm grass and molasses grass. East side remnant open forest, closed shrubland and heathland.


Target species: Vertebrates. Species detected: arboreal mammals, macropods, introduced carnivores, frogs, reptiles, birds, small mammals, rodents, echidnas

Mitigation structure/s: 9 concrete culverts purposely built for fauna, relatively uniform size (2.4 metres wide x 1.2 metres high x 18 metres long). 1.8 metres high chain-mesh fauna-exclusion fence with floppy top.

Studied the use of purpose-built culverts by fauna. Wildlife traverses were identified by sand tracking with a sand strip placed in the middle of each culvert. Tracking stations were checked every second day for eight days in spring (September 2000) and summer (December 2000), giving a total of 16 survey days. Presence of faunal species in nearby forest was determined with trapping, hair-tubes, scats, spotlighting, and frog call recordings. In addition to sand tracking during spring, wire cage and Elliott traps were set adjacent to culverts and alongside the sand strip (NB. traps were set only on one side of the sand strip possibly attracting animals across sand). In summer, hair-tubes were used and arranged adjacent to one entrance of culvert. The effectiveness of the fence was established by undertaking road-kill surveys for 20 weeks (Aug 2000 to Jan 2001).

The study identified many different species or faunal groups from the sand tracking, including arboreal mammals, macropods, introduced carnivores, frogs, reptiles, birds, small mammals, rodents and echidnas. The most commonly occurring faunal groups were bandicoots (25 per cent), rats (25 per cent), Cane Toads (14 per cent), wallabies (13 per cent) and mice (10 per cent). Results from the trapping and spotlighting confirmed that there were a number of different individuals from these fauna groups in and around the culverts. Therefore the use of the culverts by these common fauna groups may be due to high population abundance. However, the sand tracking is ineffective at distinguishing individuals and can neither accurately determine the number of individuals nor differentiate between many individuals or a few individuals crossing frequently. Two species (the Sugar Glider and the Long-nosed Potoroo) were detected in the adjoining habitat but not found using the culverts. Further study is required to determine the best mitigation measure to accommodate for these species. It is suggested that increased habitat cover surrounding the culverts may facilitate usage by the potoroo.


The authors conclude that the fence was effective in reducing mammal road-kill given that the three mammal road-kills could be accounted for by access beyond the fence and via a small breach (however there is no mention of pre-construction road-kill results so unable to distinguish if fence is effective or animals avoided crossing the road). The fence was ineffective in preventing transit by frogs as on one wet night, the road-kill survey recorded 61 amphibian carcasses. The high frog mortality and the lack of culvert use (13 frog species detected in the surrounding habitat but only two within the culvert), cause the authors to conclude that the mitigation system was inappropriate for most frogs during the study period. They suggest further assessment of more attractive culvert floors and modification to exclusion fences.

Finally, consideration will need to be given to the impact of road duplication given that there may be a distance threshold beyond which some species will not travel. This has been the case in other studies but would require more research for the species in this study.

Mansergh, I.M. & Scotts, D.J. (1989) Habitat continuity and social organization of the mountain pygmy-possum restored by tunnel, Journal of Wildlife Management, 53(3): 701-707

Location of study: Mt Higginbotham, Victoria, Australia

Habitat: Alpine vegetation, basalt scree fields, ski-slopes. The species is restricted to areas over 1430 metres elevation.

Target species: Mountain Pygmy-possum, Burramys parvus

Mitigation structure/s: Two tunnels, 0.9 x 1.2 metres in cross-section, under a road

Investigated the movement of Mountain Pygmy-possums through two tunnels built under a road specifically for the species. Tunnels were filled with rock-scree, intended to recreate the natural habitat of the possum. Rocks were also used to create a 60-m long funnel-shaped corridor on the eastern slope, between patches of habitat. The recreation of habitat meant there was no need for drift-fencing to funnel animals into the tunnel. A wire mesh grille was used at the entrances to the tunnels to avoid potential increased predation by cats and foxes.

Tunnel construction was completed by October 1985. Trapping occurred between December 1982 and 1986. Movement through one tunnel was observed using a remote-sensing camera combined with drift-fencing


(fencing in the tunnel directed the possum towards the camera). Survivorship and distribution in the area was compared before and after tunnel construction using census trapping and analysed using a Chi2 analysis.

The study suggested that fragmentation had been limiting dispersal of males, leading to lower over-winter survival of females and disrupting the social organisation of the species. After tunnel construction there was a rise in male-dispersal and female survivorship, similar to that found in less disturbed habitat.

The tunnels described appear to have been appropriately designed for pygmy-possums, with tunnel-use occurring rapidly after construction. This demonstrated the importance of utilising information about habitat requirements and social organisation in species management. In this case, dispersal was an important part of the social structure, making artificial habitat-linkages an appropriate measure. The re-creation of habitat may also have been important for the success of the tunnels.

van der Ree, R. et al (in review) Population viability analysis demonstrates success of under-road tunnel for wildlife. Ecology and Society.

Location of study: Mount Higginbotham, VIC

Habitat: High-altitude boulderfield

Target species: Mountain Pygmy-possum (Burramys parvus)

Mitigation structure/s: Two rock-lined tunnels, dimensions not given but does reference previous paper (0.9 x 1.2 metres in cross-section).

Population viability modelling of populations of Mountain Pygmy-possums (Burramys parvus) to assess the effectiveness of tunnels built specifically for the species at Mt Higginbotham. Tunnels were filled with rock scree to recreate natural habitat, as described in a previous paper (see Mansergh & Scotts 1989).

Populations of Mountain Pygmy-possum at Mt Higginbotham have been surveyed annually using a grid-arrangement of Elliot traps. Ear-tags were fitted to trapped animals. Trapping data were available from 1983 to 2003, providing information before and after tunnel construction in 1985. This also included populations 0.5 km away from the road and those likely to have been influenced by tunnel construction. These data


were used to model the population-level impact of tunnel construction and it’s effectiveness for improving population viability. Simulations were conducted over 20 years with 10,000 iterations. Parameter estimates for the model were obtained using Bayesian Markov Chain Monte Carlo methods with uninformative priors. Details of the code used are provided.

The model predicted that median female population size at the unmitigated road site was 40 per cent lower than the population unaffected by the road. The addition of tunnels reduced this, to a 15 per cent difference in population size. Minimum population size showed a similar pattern. This suggested that the tunnels mitigated the barrier effect of the road, but did not completely restore population levels. This may be due to an insufficient number of tunnels, or the impact of road mortality. Effectiveness of mitigation structures at the population level is considered important from a conservation perspective.

Hunt, A., Dickens, H. J. & Whelan, R. J. (1987) Movement of mammals through tunnels under railway lines, Australian Zoologist 24(2): 89-93.

Location of study: Maldon-Dombarton railway line, near Wollongong, NSW, Australia

Habitat: Tall open eucalypt forest interspersed with patches of sub-tropical rainforest (near culverts), and open eucalypt woodland (near tunnels). The area includes sloped areas and a plateau.

Target species: Small, native mammals

Mitigation structure/s: Five drainage culverts, from 15 x 90 cm to 240 x 300 cm (length not given). Three tunnels, approximately 3 metres in diameter and 15 – 20 metres long. Another four tunnels in the area were not monitored.

Mammals were surveyed in culverts and tunnels passing under a railway line. The culverts were intended for drainage and had been in place for decades. The tunnels were recently-built (same year as sampling) structures intended for the passage of wildlife. Surveys were conducted over 8 nights in culverts and 12 nights in tunnels. A combination of trapping (Elliot and wire cage traps), footprints, scats and hair samples were used. No statistical analyses were conducted.

The study identified some of the species using these structures but did


not look more broadly at the surrounding mammal populations or movement through structures. Three mammal species were recorded in the culverts: the Bush Rat (Rattus fuscipes), Feral Cat (Felis cattus) and Long-nosed Bandicoot (Perameles nasuta). Only rats were recorded in the smallest culverts. Mammal activity appeared less frequent in the tunnels, with feral animals predominant. Signs of dogs, foxes and Swamp Wallaby (Wallabia bicolour) were recorded. Only one unidentified track indicated the presence of small-mammals in the tunnels.

The paper admits to a limited sample size but suggests that the pattern of use is different in the two types of structures. They suggest that important factors for species use are culvert size, vegetation cover, and presence of feral animals in large tunnels. Small mammals may avoid large tunnels due to openness and predation risk. Large mammals may also be discouraged by these factors, and in addition are too large to use the small culverts. Use may also change over time. Feral animals may be more abundant soon after construction, while the area is disturbed. Vegetation growth will alter the suitability for other species. It was hypothesised that revegetation on the approaches to tunnels, and the addition of rocks and logs to act as cover, continuous through the tunnels, may assist this successional process. The study also recommends that monitoring of mitigation structures should be costed for as part of the construction process for some time after construction.

Bond, A. & Jones, D. (2008) Temporal trends in use of fauna-friendly underpasses and overpasses. Wildlife Research 35. 103-112

Location of study: Compton Road, Southern Brisbane, QLD

Habitat: Subtropical bushland, mostly dry eucalypt forest and woodland, with heath understoreys and containing lagoon systems. 324 plant species are known from the area, but no further vegetation details are given.

Target species: Terrestrial and arboreal mammals, reptiles

Mitigation structure/s: Two underpasses, 2.4 metres high x 2.5 metres wide, 48 metres long (further structural details and a photograph are provided). Both underpasses are almost identical, except for a pipe passing through the middle of one, and a grate


allowing more light into the underpass. One land bridge, hour-glass shaped with a 70 metres arc-length, base width of 20 metres and mid-width of 10 metres (further structural details and photograph provided). The land bridge includes fauna exclusion fencing continuous over the land bridge, along the side of the structure and to the opposite side of the road. The bridge also includes mulch and planted trees, shrubs and native grasses to provide cover, though these are not continuous with the surrounding forest.

Study of the use of crossing structures along a 1.3 km stretch of Compton Road, running through subtropical bushland. The section of road was upgraded from two to four lanes in 2004, at which time a variety of crossing structures were added, including underpasses, culverts with artificial ponds, a land bridge overpass, a series of glider poles, and arboreal rope-bridges connecting the tree canopy. The two underpasses and the land bridge were monitored for this study. A broader study including the population dynamics of frogs, reptiles, birds and terrestrial and arboreal mammals in the area is underway but not discussed here.

Monitoring for this study took place during two time periods. The first was six months in duration (August 2005 to February 2006) and began six months after the end of construction and the second phase was 13 months June 2006 – June 2007). In the underpasses, sand strips 1-2 cm thick, 1 m wide, and running the entire width of the raised sections were used in to detect tracks. The lowest section of the underpass was not included, as water frequently runs though here. Sand strips were placed at both ends of the underpass and checked twice-weekly during phase 1 of monitoring and once-monthly during the second phase. Smaller sand strips were also added to the underpass shelves. The land bridge was monitored using scat collection, during weekly 15-minute searches during phase 1 of monitoring and once in June 2007. Surveys of road-killed animals were undertaken once per week for four months prior to construction commencing, and recommenced immediately after construction ceased (February 2005 – June 2007).

Pearson’s correlations were used to look for changes in number of tracks and scats found each week of the study. T-tests were used to compare the use of each underpass, and whether there is an even diversity of taxa using those structures. The use of different ‘zones’ of the land bridge were also compared using ANOVA.


A wide range of species (16 taxa groups) used the underpasses and the rate of use increased during the study period. Complete crossings were identified when tracks of the same size and belonging to the same taxa were observed travelling in the same direction in both sand pads at either end of the structures. Average crossing rates were 10 per cent in Phase 1 of monitoring and 27 per cent in Phase 2. There was no significant difference detected between the level of use of the two underpasses. The most abundant tracks were from small mammals, probably rodents (probably native and exotic species). The next most abundant tracks were made by lizards. The only evidence of predation in the underpasses was a single Black Rat corpse, probably left by a Feral Cat, and a single snake-track at a time of high small mammal presence. The shelves fitted to the walls of the underpasses were also used, with 31 per cent off all tracks in the underpasses occurring on the shelves.

Scats from seven different taxa groups were found on the land bridge, with the most commonly detected scats coming from Brown Hares, with Red-necked Wallaby the next most abundant. Most scats were found on the northern slope, with few scats found on the top of the structure, thought to be because it is more exposed and lacks a clear view to shelter. Over time there was considerable variation in scat abundance, probably reflecting seasonal changes in animal activity.

Road-kill surveys show that the number of animals getting killed on the road has decreased after the exclusion fencing was installed. The large animals that collided with vehicles entered the roadway after holes had been deliberately cut in the fence by vandals.

Monitoring demonstrated that a diversity of species regularly use the underpasses and land bridge, despite the short time since construction, and disturbance still evident and plantings small. Use of structures may be increasing due to increasing familiarity and seasonal changes (cooler to warmer season).

Ball and Goldingay (2008) Can wooden poles be used to reconnect habitat for a gliding mammal? Landscape and Urban Planning 87, 140-146.

Location of study: Padaminka Nature Refuge, 14 km SW of Mackay, North Queensland


Habitat: Eucalypt woodland, dominated by Moreton Bay sh Corymbia tessellaris, bluegum Eucalyptus tereticornis and other species.

Target species: Gliding marsupials, specifically Squirrel Gliders Petaurus norfolcensis

Mitigation structure/s: Five hardwood power poles, 30 cm in diameter, extending 12 m above the ground. Poles were installed 16 – 22 m apart, traversing a 70 m clearing (agricultural land) between two forest patches. A horizontal cross-arm was installed near the top of each pole to provide a launching point, and three short lengths of PVC tube were fitted to each pole to provide a refuge to animals.

Use of the glider poles was detected using a combination of trapping in woodland at either end of the line of poles, via radiotracking, trapping on poles, releasing of animals directly onto poles and through the use of hair tubes.

Pole-released animals were capable of climbing up poles, launching from them, and landing on them after gliding. Individual Squirrel Gliders were only found using both remnants after erection of the poles, and animals were trapped on poles. One radiotracked glider denned in the pipe refuges on two poles, and glided to and from poles to forage in both remnants. Squirrel Gliders were detected in hair-tubes on 4 of the five poles.

This study confirms that Squirrel Gliders are capable of using poles to reconnect habitat and confirms that poles also have potential to provide connectivity across roads, although this has not yet been confirmed. The height and spacing of poles was identified as a critical variable to success.

Dique, D.S., Thompson, J., Preece, H.J., Penfold, G.C., de Villiers, D.L. & Leslie, R.S. (2003) Koala mortality on roads in southeast Queensland: the koala speed-zone trial, Wildlife Research 30: 419-426.

Location of study: The Koala Coast, south-east of Brisbane, Australia

Habitat: Remnant vegetation is mostly Nerang-Beenleigh open-forest alliance. The area also contains urban and industrial development, and agricultural land.

Target species: Koala, Phascolarctos cinereus


Mitigation structure/s: Differential speed signs erected every 500 m on six trial sites

Looks at the impact of differential speed signs on Koala road mortality. From August to September (previously found to be the period with highest Koala road mortality), road signs were set up displaying two different speed limits, a slower speed from 1900 to 0500 hrs, and regular speed limit at other times. Signs were placed approximately every 500 m along six trial sites. Four roads were left unsigned as control sites.

Road mortality and speed data were collected over a five-year period, from 1995 to 1999. Average annual daily traffic data came from counters and manual counts. The Moggil Koala Hospital collected data on Koala-vehicle collisions.

The differential speed signs did not significantly reduce vehicle speed or vehicle-koala collisions. Increased law enforcement may be expected to improve this situation, but is not financially viable. Factors other than traffic speed may also influence the risk of collisions and Koala mortality. Developing management strategies that address road mortality, raise community awareness and provide planning guidelines are recommended for Koala conservation in this area. Further research is required for these to be effective and practical.

Jones, M.E. (2000) Road upgrade, road mortality and remedial measures: impacts on a population of eastern quolls and Tasmanian devils, Wildlife Research 27: 289-296

Location of study: Cradle Mountain-Lake St. Clair National Park, south-west Tasmania, Australia

Habitat: Cold, wet climate, altitude 760-1080 m above sea level.

Target species: Eastern Quoll, Dasyurus viverrinus and Tasmanian Devil, Sarcophilus laniarius

Mitigation structure/s: Signs to slow speed and increase driver awareness, wildlife reflectors to deter road crossing, ramps and shelter-pipes to encourage escape from the road.


Investigates the impact of a road upgrade on populations of quolls and Tasmanian Devils. An access road was widened and sealed in June 1991. Dasyurid populations were monitored before and after the upgrade. Animals were trapped in cage traps baited with sheep heart or liver. Sixty traps were located throughout the study area and set every second month for three nights from October 1990 to April 1993. Spotlighting surveys, road kill monitoring and traffic speed measurements were also conducted along the Cradle Mountain Tourist Road.

Several measures were implemented in an attempt to reduce road mortality. ‘Slow points’ consisting of concrete barriers with a ‘give way’ sign at one end that constricted traffic to a single lane forced vehicles to reduce speed. Rumble bars were placed across the road. Speed signs and large, reflective signs were erected to raise driver awareness. Pamphlets and posters were also used to raise awareness of road mortality of wildlife. Wildlife were deterred from the road using reflectors, and encouraged to escape the road using drainage pipes and ramps.

In the 17 months following the road upgrade there was an increase in road mortality observed. The local population of Eastern Quolls (17 individuals) is thought to have become extinct, and the devil population halved in size. This is thought to be mainly due to increased traffic speed. The measures implemented to reduce mortality appear to have reduced road-kill rates to a more sustainable level. A partial recovery of the quoll population was observed, to 50 per cent of its former level. Devil populations may also have increased, though more extensive surveying is suggested to determine this.

Ramp, D. & Croft, D.B. (2006) Do wildlife warning reflectors elicit aversion in captive macropods?, Wildlife Research 33: 583-590

Location of study: A simulated road, University of New South Wales Cowan Field Station, north of Sydney, Australia (behavioural study on captive animals)

Habitat: A former orchard site, cleared of trees and shrubs to leave fairly uniform tall grasses and herbs, with some bracken (Pteridium


esculentum).

Target species: Red Kangaroo, Macropus rufus and Red-necked Wallaby Macropus rufogriseus

Mitigation structure/s: Wildlife warning reflectors

Tests the response of captive macropods to reflectors designed to deter ungulates away from roads overseas. A road was simulated by mowing a 10-m wide strip. Supplementary food was used to encourage macropods to cross from their naturally preferred side of the ‘road’. A system of lights was used to simulate headlights passing. Seventeen kangaroos and fifteen wallabies were observed using video recordings of the ‘road’. Two brands of wildlife reflectors were trialled.

The reflectors had no large effect on either species. M. rufus showed a small increase in behaviour associated with vigilance, and M. rufogriseus showed a small increase in the flight response. The small size of the behavioural responses means these reflectors are not expected to substantially mitigate animal-vehicle collisions.

Goosem, M. (2004) Linear infrastructure in the tropical rainforests of far north Queensland: mitigating impacts on fauna of roads and powerline clearings, In: Conservation of Australia’s Forest Fauna (second edition), editor Daniel Lunney, Royal Zoological Society of New South Wales, Mossman, NSW, pp. 418-434

Location of study: Wet Tropics World Heritage Area, far north Queensland, Australia

Habitat: Wet tropical rainforest, includes rugged mountain ranges

Reviews the impacts of fragmentation caused by linear infrastructure on fauna in the Wet Tropics, and techniques for mitigating these impacts.

Impacts of linear clearings in tropical rainforests

� Habitat loss and alteration

The area cleared for linear infrastructure is smaller than, for example, clearing for forestry or rural or urban development. However, the


impact of disturbance also extends into the forest. Roads constructed near streams can alter habitat for a large distance, including changes to erosion, sedimentation and flow, and pollution from highway runoff. Maintaining canopy cover over the road surface can reduce erosion.

� Disturbance effects - the emission of matter and energy

Linear infrastructure can create disturbance in adjacent areas, including noise, headlights, vibration, movement, electromagnetic radiation and pollution. Many animals avoid areas near roads, so reducing the amount of suitable habitat. Noise can cause stress and hearing damage, alter behaviour and disrupt communication. The penetration of noise is altered by vehicle type and speed, and road topography (e.g., noise travels further on steep slopes). Dense rainforest vegetation can help to screen particulates from penetrating the forest edge. Heavy metal accumulates along road edges, and may enter the food chain via soil invertebrates.

� Edge effect

Abrupt margins between relatively natural habitat and clearings cause alterations in microclimate, vegetation structure and floristics, so altering habitat for fauna. Changes include increased wind speed and turbulence, and increased light penetration. This in turn leads to greater soil temperature fluctuations, higher evaporation, and decreased humidity and soil moisture. Higher light levels favour disturbance-adapted species, including weeds and woody vines. Winds and vines can cause increased canopy damage and tree death. Rainforest specialist species avoid these areas, and there is a rise in gap and edge specialist birds and other generalist species. Restoration of rainforest across power line clearings can decrease the extent of microclimate alteration.

� Spread of weeds, feral animals and fauna from other habitats

Clearing for linear infrastructure creates disturbance that facilitates weed invasion by reducing competition from native species. Pests and exotic fauna are also more able to penetrate forests from cleared areas. Weeds can impair ecosystem function and limit the recruitment of native species. Restoration under powerlines including weed control can allow the re-establishment of rainforest vegetation, and the return of


rainforest fauna. Feral species (e.g., cats, dogs, Cane Toads) may use roads for movement and hunting.

� Road mortality

Road mortality of small animals is difficult to survey, meaning the magnitude of this impact poorly known. The impact can be significant, and has been linked to local extinctions of some species. A population sink occurs when road mortality exceeds recruitment. Certain behaviours can cause species to be prone to road mortality, for example mass migrations or dispersal. This makes seasonal factors important. Other important factors include clearing width, presence of gullies or creeks, and traffic speed. It is thought that wide roads can reduce mortality by increasing avoidance. Embankments can also create barriers to crossing. Riparian corridors may ‘funnel’ animals towards the road.

� Linear barrier effects

The combination of the impacts of linear infrastructure means they create barriers for many species. This effect is increased by fences, concrete dividers, and other physical barriers. Complete subdivision of a population can lead to the remaining populations being smaller, less viable, and more vulnerable to local extinction. In the long term inbreeding may also occur. The degree of the barrier effect depends on clearing width, traffic volume and speed, and whether canopy connections exist.

Mitigation and management of linear clearing impacts for rainforest fauna

� Reducing habitat loss - rainforest avoidance and rehabilitation of unused clearings

Avoiding sensitive habitat or not building roads is often an impractical mitigation technique as human populations expand have increased need for transport, energy and water infrastructure. Rehabilitation of old roads and clearings is costly, and passive regeneration by succession is slowed or stopped by the presence of grasses and woody weeds and increased fire frequency. Many rainforest species have difficulty colonising compacted road surfaces.


� Maintaining canopy connectivity

Retention of tree canopy is thought to ameliorate many impacts of linear clearings. This includes reducing changes in microclimate, vegetation structure and composition, and fauna composition. It is also thought that this reduces the barrier effect for many species. Erosion, weed invasion, and feral species are also reduced in the process. Economic gains are expected from reduced road-verge maintenance, and also reduced pollutants due to lack of herbicide spraying. Safety issues related to falling branches mean canopy connectivity cannot be maintained on major, high-speed roads. It is also unfeasible above powerlines.

� Minimising clearing width or mortality?

There is a trade-off between mortality risk and the barrier effect, related to road with. Small widths and minimal clearing maximise remaining habitat and canopy cover, but can also increase road mortality. On most low use, narrow roads minimising clearing width is thought to provide a net positive impact due to increased connectivity, with mortality-risk not expected to be a problem for common species. Areas where this may not be the case include high speed/high traffic roads, where rare or threatened species are involved, or where a life-cycle phase is particularly vulnerable to road mortality (e.g. where the road is attractive for warmth, foraging, breeding or for dispersal or migration).

� Fauna-sensitive design - engineering options

Bridges that allow canopy to remain underneath are recommended for movement of canopy, understorey and ground-dwelling fauna. This should reduce mortality and barrier effects. Noise and other disturbance may still limit crossing by disturbance-sensitive species. Underpasses are a cheaper alternative, and can include refuge from predators in the form of ropes and branches for arboreal species, and rocks and logs for ground-dwellers. Use is expected to increase over time as nearby revegetation matures. Fencing may also be required to prevent animals crossing the road, and must be designed appropriately for the target species. For example, fences need to continue underground for


burrowing species; a solid section at the base can stop smaller species from passing through. Fencing can also funnel fauna towards crossing structures. Culverts designed for intermittent water flow may also be used by fauna as crossings. Adding ledges above stream level may create dry passages for small species. Rope tunnel overpasses may also facilitate the movement of arboreal species where no canopy connections exist. The effectiveness of these on large roads (e.g. 2-4 lane highways) is unknown.

Reduced traffic speed could reduce road mortality. Signage and other measures including rumble strips and bicoloured tarmac, to give the impression of a narrower road, have not been found to have a lasting impact on reducing speed.

The impact of power line clearings may be reduced by building tall towers that allow powerlines to swing above the tree canopy. Clearing is then minimised to the area of the tower, and revegetation of low-growing species can occur here. Using helicopter-landing pads for maintenance of towers also reduces the need for clearing.

8.2 Australian Reports, Theses and Conference Proceedings

Abson, R. & Lawrence, R. E. Slaty Creek wildlife underpass study, Final Report (Centre for sustainable regional communities, Latrobe University, Bendigo, Kyneton, 2003).

Abson, R. & Lawrence, R. E. in Proceedings of the 2003 International Conference on Ecology and Transportation (eds. Irwin, C. L., Garret, P. & McDermott, K. P.) 303-308 (Center for Transportation and Environment, North Carolina State University, Lake Placid, New York, USA, 2003).

Location of study: Calder Freeway, Central Victoria

Habitat: Black Forest. Eucalyptus dominated forest

Target species: Birds, arboreal mammals, frogs and reptiles

Mitigation structure/s: One underpass (extended bridge, 12 metres high, 70 metres wide, and 100 metres gap between forest on either side of road) and six culverts (size not given).

Investigated the occurrence of all wildlife within the underpass and in the forest to the east and west of the bridge. The underpass was designed to be larger than required for water flow along Slaty Creek to


provide for the movement of wildlife. In addition to the single large bridge, 6 culverts of unreported size were also installed near by. It is unclear if these culverts were designed primarily for wildlife, water flow or both.

The bridge and freeway were constructed from 1997 – 1999, and surveys of use were conducted for 4 days per month from July 2002 – June 2003. The extended bridge cost ~AUD3 million to build, and monitoring cost ~AUD70,000. A large range of survey techniques were used to detect the presence of species, including active searches, electronic bat detectors, bird surveys, trapping and hair tubing, spotlighting, pitfall trapping, scat collection and sand tracking pads. Surveys were conducted over a similar survey area (~50 metres x 50 m) within the underpass, as well as in forest on both sides of the road. Culverts were surveyed by sand tracking, with a single culvert also being surveyed with a hair tube.

Thirty one species of mammal (including eight introduced species), four species of reptile, six species of frog and 37 species of bird (one introduced species) were detected within the underpass during the study. The culverts were used by a smaller number of species of mammal, reptile and amphibian. Concurrent surveys of wildlife killed along the fenced section of the Calder Freeway include 14 species of mammal (6 introduced species), one amphibian and 11 species of bird.

The study concluded that the underpass with retained vegetation (considerable effort during and after construction to protect existing plants), and extensive revegetation works under the raised bridge structures was capable of supporting a comparable number of species of mammal, bird, reptile and amphibian as the forest immediately adjacent to the Calder Freeway. A smaller objective of the study was to compare the use of culverts and an underpass to allow local vehicles access under the freeway. Fewer species were detected in these structures, and this was explained by a reduction in habitat suitability for non-detected species within the wildlife crossing structures.

The study focussed on determining the presence of wildlife in forest adjacent to the road, and within the wildlife crossing structures. It was implied that if wildlife were detected within the structures that they would be capable of using it to cross from one side to the other. This is likely, however actual crossing was not shown.

Suggestions for future work and research focus on protecting and re-


establishing vegetation under the bridge, increasing the depth and diversity of the leaf litter layer, modifying the bridge and piers to provide roosting habitat for bats and the trialling of rope bridges and glider poles for gliding mammals.

Goosem, M. (2005) Effectiveness of rope bridge arboreal overpasses and faunal underpasses in providing connectivity for rainforest fauna, International Conference on Ecology & Transportation 2005 Proceedings, pp. 304-322.

Investigated the use of faunal underpasses and overpasses in highland rainforest of northeast Queensland. The crossing structures were intended specifically for wildlife.

The report admits that usage data do not prove that crossings are sufficient to maintain populations on either side of the road. The importance of population-level studies is acknowledged, but this requires long-term monitoring and funding. Usage and mortality data are considered a cheaper indication of effectiveness, especially if long-term studies can be conducted.

a) Faunal underpasses

Location of study: Wet Tropics World Heritage Area, northeast Queensland

Habitat: Highland rainforest habitat, at approximately 1,100 metres altitude.

Target species: Designed for multiple-species use, but focused specifically on the rare species Lumholtz’s Tree-kangaroo and the Southern Cassowary.

Mitigation structure/s: Four underpasses, consisting of galvanised steel arches with a concrete base, 3.4 high x 3.7 metres wide. The height was chosen to allow cassowaries (1.5-2 metres high) to pass through, and provide line-of-sight to surrounding rainforest. The floor of the underpasses were covered in soil, leaf and branch litter, rocks and logs. ‘Furniture’ including large poles was also included, and a rope swung between hooks on the ceiling. Enough space for larger animals to pass through was maintained through the centre of the underpass. Underpass placement was as close to rainforest remnants as possible, and rainforest trees including food plants were planted to create corridors between remnants.


Four fauna underpasses were included as part of a road upgrade in 2001, which involved straightening and widening a highway from two to four lanes. Underpasses were monitored using sand track pads. Road-kill data were also collected, and some trapping for small mammals and bird surveys were conducted in the area.

Road transects 0.5 km in length were surveyed for dead animals once per week for 12 months prior to construction. These surveys continued in a similar way after construction using 0.5 km transects along similar habitat 5 km to the north of the crossing structures. Small mammal trapping was conducted before construction between July and November 2000. Trapping was conducted monthly over 3-4 consecutive nights. 5-m grids of 20 Elliot traps and two cage traps were used, replicated two times in each of four different habitats. The habitats surveyed were abandoned pasture, pasture over-run by Lantana camera, rainforest on the edge of the north and south main rainforest areas, and interior rainforest (more than 100 metres from the edge). Three years after construction, small mammal trapping was conducted in May-August 2004. Three replicates were used in each habitat type. Sub-sites were also looked at, with separate monitoring of north and south sites. Birds were point censused at nine sites in September 2004. Two visits were conducted per site (morning and evening).

Underpasses were monitored using sand track beds and infra-red triggered cameras. Sand beds consisted of a 1-m wide strip of sand, 5 cm deep and running the entire width of the underpass. One sand bed was placed in the centre of each underpass, and monitored weekly.

Road-kill data were investigated using a Chi2 analysis to compare data on separate years, highways, vertebrate groups and habitat guilds. Kruskal-Wallis nonparametric analysis of variance was used to examine the abundance of small mammal species, and compared using Mann-Whitney U-tests. Bird groups were compared using ANOVA. Underpass use was divided into sampling periods by season, and temporal variation examined using homogeneity tests. Relationships between use by feral and native species was looked at using correlations.

The most common groups using the underpasses were bandicoots, Red-legged Pademelons and brushtail possums. Of the target species, Lumholtz’s Tree Kangaroo was recorded using the underpasses on two occasions. There were no records of Southern Cassowary, possibly because it has become very rare in the area.


Possible reasons suggested for some species not using underpasses include a lack of time for habituation to structures and establishment of corridor vegetation, failure to detect all species using underpasses, corridors may not be wide enough and only provide edge habitat, there may be insufficient cover within underpasses, and animals may be deterred by traffic noise or other disturbances.

The strategy of using a large underpass that incorporated simulated forest-floor and a range of cover and escape options for smaller species was considered a success, as mammals and birds of a range of sizes and levels of ‘shyness’ were detected using underpasses. Conditions suitable for some species may still not have been met, e.g. arboreal possums.

Occasional use by feral predators did not appear to deter use by native species. The inclusion of cover and poles may have helped this. There was only one observation of predation occurring near an underpass.

b) Fauna overpasses

Location of study: Wet Tropics World Heritage Area, northeast Queensland. The rope tunnel was located 30 km southwest of Cairns.

Habitat: Highland rainforest

Target species: Arboreal mammals, particularly rainforest ringtail possums including the Lemoriod Ringtail Possum (Hemibeldeus lemuroides), the Herbert River Ringtail Possum (Pseudochirulus herbetensis) and the Green Ringtail Possum (Pseudochirops archeri).

Mitigation structure/s: One rope tunnel, 50 x 50 cm deep made of 10 mm silver rope held taut with plastic spaces and attached to wooden poles in the trees at the road-edge. The entire structure is 7m tall and 14 metres wide. Three rope ladder arboreal overpasses, resembling a rope-ladder. Two were installed over a forestry track and were 10 metres long and 5 metres above ground-level. The span of these bridges were 5 and 7 metres respectively. One was 50 cm wide, the other half-width (25 cm). The third overpass was installed over a 2-lane tourist road, with a 14 metres gap between the trees.

The rope tunnel overpass was constructed in 1995, but no monitoring was conducted until 2000. A single-strand of rope was also erected in 2000. Monitoring was conducted using a 1 metres wide net of fine mesh under the rope tunnel to intercept dropped scats. Scats were collected several days each month from January to October, 2000, then


permanently from August to October, 2001. Hair samples were collected using double-sided tape fixed around the rope. An infra-red triggered camera was periodically installed inside the rope tunnel from January to October 2000. Spotlighting of the structure was conducted for 40 hours over 10 randomly selected nights from July 2000 to Feb 2002. This included nights in both wet and dry seasons.

At the rope-ladder overpasses, monitoring was conducted using spotlighting for 40 hours at the forestry track and 70 hours at the tourist road. Scat collection using funnels of wire mesh going to a PVC pipe collector was conducted between September and December 2001. Hair samples were collected from August to November 2001 using a curtain of wire frame 55 cm in diameter draped with double-sided tape and attached centrally to the bridge.

The rope tunnel was used by all target species. Results on the single rope-strand were inconclusive, with no photographic evidence obtained. Natural canopy connections were preferred to artificial structures when these were available. Where these were not available, the rope ladder was used by the majority of species.

At the 14-m rope-ladder overpass the first photograph of a possum using the structure occurred after 5 months. The first recorded crossing occurred six weeks after this. Usage then rose to a level of approximately 1 crossing per hour. This suggests a habituation period may have occurred.

Australian Museum Business Services Consulting (January 2001) An investigation of the use of road overpass structures by arboreal marsupials, Roads and Traffic Authority Final Report

a) Movement of gliders

Location of study: Termeil State Forest, south coast New South Wales.

Habitat: Open forest dominated by red bloodwood (Corymbia gummifera), blackbutt (Eucalyptus pilularis), spotted gum (Corymbia maculata) and turpentine (Syncarpia glomulifera). Understorey dominated by hairpin banksia (Banksia spinulosa) and mountain devil (Lambertia formosa). Altitude 50 – 70 metres ASL.

Target species: Sugar Gliders (Petaurus breviceps), Yellow-bellied Gliders (Petaurus australis), and Greater Gliders (Petauroides volans).


Mitigation structure/s: A single glider pole, consisting of a 12 metres timber pole with horizontal cross member to act as a guiding platform, with aluminium ‘predator shield’ (schematic available).

Investigated the movement of gliders on both sides of a three-lane section of the Princes Highway, narrowing to two lanes to the north and east of the study site. A single gliding pole was installed, intended to act as an overpass by reducing the distance required to glide across the road. The structure was located 20 m from the western and 40 m from the eastern forest edge. It was fitted with a movement/heat sensor to monitor use by gliders over a 17-week period (June 23rd to November 19th, 2000).

Gliders in the area were surveyed and tracked using a combination of trapping (mark-recapture), incidental observations, radio-tracking, and spotlighting. Surveys occurred from May-September 1998, and again in June 1999, December 1999 and November 2000. Trapping occurred monthly during these periods on both sides of the highway, using a combination of Elliot and cage traps. Greater Gliders were recorded from spotlighting and incidental observations. Captured animals were marked using ear-notches to allow individuals to be identified. Nine Sugar Gliders, two Yellow Bellied Gliders and one Greater Glider were fitted with radio-transmitters and tracked during the study period. Spool and line tracking was attempted in August 1999, but was abandoned because it was considered too high a risk to the animals and the thread was too high in the canopy to be easily detected.

Of the Sugar Gliders tracked, six regularly crossed the highway. It is assumed they crossed by gliding, though this was only observed once. None of the Yellow-bellied or Greater Gliders were found to cross the highway. The sensor did not record any animals using the gliding pole.

It was suggested that design features of the pole (e.g. height, configuration, vegetative cover around the pole) and its placement need to be re-examined and trialled. It is also possible that the sensor failed to record animals. Retaining vegetation in median strips of multi-lane roads is recommended.

b) Movement of Common Ringtail Possums

Location of study: Wakehurst Parkway, Allambie Heights, Sydney

Habitat: Sydney sandstone ridgetop complex. Dense, tall heathland on the eastern half of the study site, dominated by old an banksia (Banksia


serrata), heath banksia (Banksia ericifolia), Sydney golden wattle (Acacia longifolia), forest oak (Allocasuarina littoralis), needlebush (Hakea sericea), and dagger hakea (H. tereticornis). Taller vegetation (8-10 m) on the eastern side, dominated by grey gum and silvertop ash (Eucalyptus sieberi) with dense tick bush (Kunzea ambigua) dominating in many areas, with scattered broad-leafed hakea (Hakea dactyloides), heath banksa and dwarf apple (Angophora hispida).

Target species: Common Ringtail Possum (Pseudocheirus peregrinus).

Mitigation structure/s: One overpass consisting of a wire cable strung between two poles on either side of the road. An aluminium, alloy perforated tube was suspended around the wire cable for protection of animals (schematic available).

Investigated the movement of ringtail possums across the two-lane Wakehurst Parkway (see Fig. 4) via the overpass specifically installed for the movement of possums. The structure was monitored over an 8-month period using a movement/heat sensor and data-logger installed within the tubing (From January 2000). Scats in the area were also noted. Ringtails in the area were captured by hand. Four animals were fitted with radio-collars and tracked for the first few hours after sunset over eleven months (March 1999-Febuary 2000). It is unclear if this canopy bridge / overpass was constructed to mitigate a demonstrated barrier effect of Wakehurst Parkway or if it was an experimental trial.

Animals began using the overpass soon after construction (January 28th 2000), with nocturnal records and scat observations suggesting Common Ringtail Possums were using the structure. The number and frequency of complete crossings could not be determined from the survey method. Radio-tracking of ringtail possums was also attempted, but it gave inconclusive results due to poor reception and interference. Animals were known to cross the road due to incidental observations. In addition, eight ringtails and one Common Brushtail Possum (Trichosurus vulpecula) were found killed on the road during the study.

It was assumed that Common Ringtail Possums used the structure to cross the road, although the monitoring used could not distinguish individuals or confirm that they made a complete crossing. Changes to monitoring, possibly including the use of infra-red cameras, were recommended. Fencing to funnel animals to overpasses was suggested as potentially useful for reducing road-mortality, though it was not tried in this study.


Australian Museum Business Services Consulting (June 2006) Aninvestigation of the use of the road overpass at Wakehurst Parkway by arboreal mammals, NSW Roads and Traffic Authority

Location of study: Wakehurst Parkway, Allambie Heights, Sydney

Habitat: Sydney sandstone ridgetop complex. Dense, tall heathland on the eastern half of the study site, dominated by old man banksia (Banksia serrata), heath banksia (Banksia ericifolia), Sydney golden wattle (Acacia longifolia), forest oak (Allocasuarina littoralis), needlebush (Hakea sericea), and dagger hakea (H. tereticornis). Taller vegetation (8-10 m) on the eastern side, dominated by grey gum and silvertop ash (Eucalyptus sieberi) with dense tick bush (Kunzea ambigua) dominating in many areas, with scattered broad-leafed hakea (Hakea dactyloides), heath banksa and dwarf apple (Angophorahispida).

Target species: Common Ringtail Possum (Pseudocheirus peregrinus).

Mitigation structure/s: One overpass consisting of a wire cable strung between two poles on either side of the road. An aluminium, alloy perforated tube was suspended around the wire cable for protection of animals (schematic available).

Investigated a single overpass for possums across Wakehurst Parkway (described in further detail in AMBS 2001). The 2006 report included a mark-recapture study to determine crossings, road-kill survey, and cameras with infra-red sensors set up at each end of the overpass. Cameras were installed November 7th 2005 and left in place for 108 nights of monitoring. They were checked and maintained weekly. Ringtail and brushtail possums were captured from either side of the road during a three-day period throughout the study area, then weekly along transects either side of the overpass. Possums were collared to allow individuals to be identified if photographed using the overpass. Weekly road-kill surveys were conducted on either side of the overpass, and other road-kill data were obtained from local organisations (e.g. veterinary practices, council maintenance crews).

A single Common Brushtail Possum (Trichosurus vulpecula) was photographed using the structure during the monitoring period, but failed to make a complete crossing and was only on the wire for a few seconds. Despite the use of exclusion fencing, seventeen animals of


eight species were recorded killed on the road. The most common species were the Common Brushtail Possum and Common Ringtail Possum. Six arboreal mammals were captured during the mark-recapture study. None of these were recorded using the overpass, or crossing the road elsewhere, and there was a high rate of recapture of collared animals.

Previous monitoring of this overpass (see AMBS 2001 report) demonstrated that the structure was appropriate for ringtail possums. Use was also expected to increase over time. It is possible that frequent use by one individual, such as a dominant male, has deterred other individuals, but this is considered unlikely. Another possibility is that vegetation clearing around the base of the western pole during construction of exclusion fencing may have made animals reluctant to use the structure.

Replanting in the area is recommended, and the inclusion of ropes between the pole and surrounding vegetation in the short-term. Vegetative cover at entrances of overpass structures is considered important for future design and maintenance of similar structures. The addition of exclusion fencing needs to be carefully considered, taking into account how much habitat removal is required to install and maintain the fence, and how the fence leads animals to migration structures. Incorporating existing trees in structure design where possible is also suggested.

Australian Museum Business Services Consulting (June 2002) Fauna underpass monitoring, Stage One - Final Report - Taree, Roads and Traffic Authority

Location of study: Pacific Highway, Taree, NSW

Habitat: Sparse Acacia and Casuarina saplings, tall grasses, rushes and shrubs close to underpasses. Further away vegetation included tall open forest, medium swamp forest or open paddocks, with low casuarina forest near creeks and mangroves. Rainfall data are also provided.

Target species: Small to medium-sized mammals, macropods, arboreal mammals

Mitigation structure/s: Two bridge fauna underpasses, one steel arch and four other fauna underpass/drainage structures. Dimensions were not reported, but were considered as part of the analysis.


This report contains monitoring at Taree during Stage 1 of a larger project spanning several reports. Monitoring here was split into two Episodes: 17 August to 2 September 2000 (Episode 1) and 10-26 November 2000 (Episode 2). A range of crossing structures along the Pacific Highway were monitored, including some specifically designed for fauna and some that also serve as drainage structures.

Three sand trays were set up per underpass to record animal tracks, one at each end and one in the middle. These were 1 m wide and extended the width of the underpass. One underpass also had sand spread around access points to furniture (logs) to detect animals climbing the poles. Sand trays were checked every four days over a twenty-day period. The entire underpass and a 20 m radius around the entrance was also checked for signs such as scats and further tracks. Road-kills nearby were also recorded as well as the condition of the underpass and fencing.

The most frequent species groups recorded using the underpass were Fox (Vulpes vulpes), antechinus and rat. Bandicoots were also common in Episode 2. Note that tracks often cannot be identified to species-level, necessitating the use of broader groupings. Most fauna appeared to make a complete passage through the underpass. Narrow culverts appear to deter wallabies and possibly some other species. Presence of stumps, length of the culvert/tunnel and the small diameter of some culverts may influence behaviour of some animals. This part of the study also led to some recommendations for improving the monitoring regime in other areas, including the inclusion of some underpasses for photographic monitoring in Stage 2.

Australian Museum Business Services Consulting (June 2002) Fauna underpass monitoring, Stage One - Final Report - Bulahdelah to Coolongolook, Roads and Traffic Authority

Location of study: Pacific Highway, Bulahdelah to Coolongolook, NSW

Habitat: Habitats include dry and wet sclerophyll forest, temperate forest, swamp forest, and open grassland. Steep topography. Rainfall data are given.

Target species: Small to medium-sized mammals, macropods, arboreal mammals.

Mitigation structure/s: Seventeen 3 x 3 m box culverts, three


precast concrete arches and two multi-span bridges. One of the culverts contains a trial skylight.

This report contains monitoring from Bulahdelah to Coolongolook during Stage 1 of a larger project spanning several reports. Monitoring here was split into two Episodes: 17 August to 2 September 2000 (Episode 1) and 10-26 November 2000 (Episode 2). A range of crossing structures along the Pacific Highway were monitored, including some specifically designed for fauna and some that also serve as drainage structures. One was specifically aimed at allowing frogs to cross.

Survey method as given in AMBS 2002 (Taree report).

Bandicoots were the most frequent species group recorded using the culverts, followed by fox and cat. The majority of records indicated a complete passage of the underpass. Stumps blocking the entrance may deter larger species (e.g. wallabies).

Recommends underpasses for future photographic monitoring. Log removal or size-reduction is also recommended to allow greater access by fauna to some underpass entrances.

Australian Museum Business Services Consulting (June 2002) Fauna underpass monitoring, Stage One - Final Report - Brunswick Heads, Roads and Traffic Authority

Location of study: Pacific Highway, Brunswick Heads Bypass, NE NSW

Habitat: Occurs at the overlap of tropical and temperate zones, with high rainfall and rich soils. This leads to a high diversity of habitats. Specific information on vegetation in the area is not given. Rainfall data are reported.

Target species: Small to medium-sized mammals, macropods, arboreal mammals, monotremes, rodents, reptiles, frogs.

Mitigation structure/s: Two bridge underpasses and eleven box culvert underpasses, either 2.4 x 1.2 or 2.4 x 1.5 m in size. Lengths are not reported, but are discussed.

This report contains monitoring from Brunswick Heads Bypass, running parallel to the Pacific Highway, during Stage 1 of a larger project spanning several reports. Monitoring here was split into two Episodes: 17 August to 2 September 2000 (Episode 1) and 10-26 November 2000 (Episode 2). A range of crossing structures were monitored, including


some specifically designed for fauna and some that also serve as drainage structures.

Sand trays were set up as previously described (see AMBS 2002 Taree report). These were monitored every four days during Episode 1 and every two days during Episode 2. Recognisable tracks and direction of movement were recorded. Other evidence of animals in the area was also recorded (see details in AMBS 2002 Taree report). Data on species thought to occur in the area were obtained from Australian Museum and NPWS databases. No detailed statistical analyses were conducted. Vegetation and location data were not collected in an easily comparable way.

The most frequent species groups using the underpasses were large wallabies and bandicoots. All underpasses had a higher proportion of native animals crossing. Some threatened species were recorded: Koala, and possible Long-nosed Potoroo and Black Bittern. A greater number of complete passages were made through bridge underpasses than box culverts. Road-kills during Episode 2 included bandicoots, rats, Swamp Wallaby, and Common Ringtail Possum. Some breaches may be present in the exclusion fencing. Little difference was seen between underpasses opening into cleared pasture, native vegetation, or flooded areas. It is thought that short underpasses encouraged usage by fauna.

The mitigation measures along the bypass provided opportunities for a large number of species to cross. It is thought to provide more opportunities than the other Pacific Highway sites monitored due to higher quality, less disturbed habitat that continues close to underpass entrances compared to other sites. Continuous habitat exists on either side of the bypass. Construction was also thought to have kept disturbance to a minimum. Future monitoring using cameras at underpasses is suggested.

Australian Museum Business Services Consulting (March 2001) Fauna underpass monitoring, Stage One - Final Report - Herons Creek, Roads and Traffic Authority

Location of study: Pacific Highway, Herons Creek, NSW

Habitat: Closest habitat consisted of lantana, tall grasses, shrubs and creeks. The closest pre-highway vegetation varied from medium open regrowth forest to tall open forest.


Target species: Macropods, small mammals, reptiles, rodents

Mitigation structure/s: One bridge fauna underpass, one pipe culvert and one box culvert

This report contains monitoring from Herons Creek during Stage 1 of a larger project spanning several reports. Monitoring here was split into two Episodes: 17 August to 2 September 2000 (Episode 1) and 10-26 November 2000 (Episode 2). A range of crossing structures along the Pacific Highway were monitored, including some specifically designed for fauna and some that also serve as drainage structures.

Sand trays were set up and further signs of animals using the underpasses were recorded as previously described (see above reports). Data on animals thought to occur in the area were obtained from previous surveys and databases. No comprehensive statistics could be used due to a lack of replication.

Rats were the most frequent group using underpasses, followed by bandicoots, water dragons and macropods. The rats could not be identified to species from their tracks but were assumed to be native. During Episode 2, only native species were recorded. One Swamp Wallaby was recorded killed on the road in the area during Episode 2. It is thought that narrow culverts deterred wallabies and some other species. Fauna exclusion fencing appeared intact and is thought to be working effectively. The area is not recommended for additional monitoring, though a system to correctly identify rats using the area is suggested.

Australian Museum Business Services Consulting (June 2001) Pacific Highway - Fauna Underpass Monitoring, Stage Two, Episode Three Taree, Roads and Traffic Authority

Location of study: Pacific Highway, Taree, NSW

Habitat: Sparse Acacia Casuarina saplings, tall grasses, rushes, shrubs and water. Further away from underpass habitat varied from tall open forest, medium swamp forest to open paddocks and low casuarina forest near creeks and mangroves.

Target species: Small to medium-sized mammals, macropods, arboreal mammals.

Mitigation structure/s: One steel arch, a single-span bridge and one


drainage pipe culvert, either 1350, 1800 or 2100 mm in diameter.

This report contains monitoring from Taree during Stage Two, Episode Three of a larger project spanning several reports. Stage Two involved further monitoring of underpasses surveyed using sand track pads in Stage One (see previous report) and consisted of Episodes 3, 4 and 5. Monitoring for Episode Three occurred from 14 February to 14 March, 2001.

Three underpasses were chosen based on security, lack of flooding and relatively high use by fauna (determined during Stage One). Cameras connected to an infra-read beam trigger running the width of the underpass were installed. Sand-tray monitoring was also used simultaneously, and checked every 7 days over a 35 day period.

The most frequent species recorded in the area during Episode Three were Long-nosed Bandicoot, Black Rat (Rattus rattus), and Brown Antechinus. Other species photographed included water dragon and House Mouse. One set of fox prints was recorded, the only evidence of introduced predators. Most animals appeared to make a complete crossing. Flooding during the study caused significant data loss, and prevented usual movement patterns. Brushtail possum, hares and rabbits were recorded killed on the road during the Episode.

Australian Museum Business Services Consulting (June 2001) Pacific Highway - Fauna Underpass Monitoring, Stage Two, Episode Three, Bulahdelah to Coolongolook, Roads and Traffic Authority

Location of study: Pacific Highway, Bulahdelah to Coolongolook, NSW

Habitat: Not described (but see previous reports)

Target species: Small mammals, macropods, rodents, introduced predators, reptiles

Mitigation structure/s: Eleven underpasses. Included 3 x 3 m box culverts.

This report contains monitoring from Bulahdelah to Coolongolook during Stage Two, Episode Three of a larger project spanning several reports. Stage Two involved further monitoring of underpasses surveyed using sand track pads in Stage One (see previous report) and consisted of Episodes 3, 4 and 5. Monitoring for Episode Three occurred from 14


February to 14 March, 2001.

Eleven underpasses were monitored during Episode Three using infra-red triggered cameras in combination with sand track pads as previously described (see above reports).

Bandicoots were the most common species using the underpasses in Episode Three. Long-nosed Bandicoots (Perameles nasuta) are shown in the example photographs, it is not clear whether Northern Brown Bandicoots (Isoodon macrourus) were also recorded. No clear difference between underpasses with and without skylights was observed.

Possible factors identified for increasing use by fauna include adjoining habitat and its degree of disturbance, underpass length, distance to pre-highway vegetation, steepness of approaches, visibility of entrance from adjoining habitat, topographic position (e.g., gully or ridge), and unobstructed view of horizon or adjoining habitat from within the underpass. Interactions may also be important. None of these factors are analysed as part of the report, though relationships to vegetation type will be analysed in future reports. Flooding and wet conditions resulted in some data loss and restricted movement patterns.

The point is made that the number of passages may not be the best measure of underpass success or usefulness. The range of species and type of use are also important (e.g., dispersal by rare species).

Australian Museum Business Services Consulting (June 2002) Fauna underpass monitoring Stage two, Episode five, Bulahdelah to Coolongolook, NSW Roads and Traffic Authority

Location of study: Pacific Hwy, Bulahdelah to Coolongolook, NSW

Habitat: Dry sclerophyll forest, wet sclerophyll forest, temperate rainforest and open grasslands (dominant species given). Presence of vegetation types and topographical features near each culvert are described.

Target species: Predominantly mammals, some birds, frogs and reptiles.

Mitigation structure/s: Eleven 3 x 3 m concrete box culverts. The length of culverts for the broader study as a whole varied from 37 to 42 m, with one significantly longer at 52 m. One culvert was a 3 x 3 m twin-cell box culvert with a skylight in the roof of one side only. All


eleven structures had log and pole furniture included. Most had a mulch floor, two had a concrete floor, and two a combination of dirt and mulch. All had exclusion fencing to keep animals off the road, most of which was continuous. The twin culvert included fencing running 1000 m to the north and 200 m south.

This report contains monitoring from Bulahdelah to Coolongolook during Stage Two, Episode Five of a larger project spanning several reports. Monitoring for Episode Five occurred from 11 November 2001 to 16 December 2001. This report also included broader information on the study as a whole, which involved monitoring the use of culverts by a variety of species under a 23-km stretch of the Pacific Highway. The entire study took place over 18 months (July 2000 - December 2001).

Species were detected using a combination of sand trays and cameras linked to infra-red beam triggers. Three sand trays were placed in each culvert, one at each entrance and one at the centre, to indicate complete or partial passage. These were checked weekly and brushed smooth. Cameras were present for the entire five weeks. Observations were compared to previous information on the species present in the area. Road-kill data were also collected from the study area.

Only mammals and some reptiles were identified to species level in this study. Microbats could also not be identified to species level. At least 19 native species and 4 introduced species of mammals were recorded using the underpasses. Bandicoots (including both Long-nosed and Northern Brown Bandicoots), Goannas (Varanus varius) and cats were the most common fauna recorded. Eleven animals found killed on the road during the study. This included a Koala (Phascolarctos cinereus), Red-necked Wallabies, Eastern Grey Kangaroo, ringtail possum, bandicoot (species not given), Rat (Rattus sp.), tortoise (species not given) and kookaburra. The majority of these occurred where there is a break in fauna-exclusion fencing.

Foxes and wallabies had significantly higher passage frequency when the underpass entrance was more than 50 m from surrounding dry sclerophyll forest. Brushtail possums crossed more frequently when there was less than 50 m between the entrance and vegetation. Other species did not show a significant difference.

Wallabies and Red-necked Pademelons were more frequent in the longest underpass, though this may be due more to differences in vegetation. Cats were more common in shorter underpasses. No


evidence of hunting behaviour or higher predator concentration at underpasses was observed, though detailed analysis was beyond the scope of the study. No significant differences were detected between the frequency of passages through the sides of twin-box culvert with and without a skylight. Scratches and scats suggest that the logs present are used by both native and exotic animals, but this could not be analysed in this study. The species present are known to make complete passages of culverts in the absence of log and pole furniture.

Fauna surveys in the surrounding area are recommended concurrent with monitoring to assist in interpreting results. This may not be required if sufficient information is already available. The study recommends that where possible vegetation should not be cleared between the underpass entrance and surrounding vegetation. Less than 50 m clearing is likely to be preferred by arboreal or smaller species, larger than 50 m may be appropriate for species attracted to cleared areas for foraging (e.g., wallabies). Skylights are not thought to be required for the passage of nocturnal fauna for the underpass lengths investigated here. The importance of ongoing maintenance of exclusion fencing is highlighted, and fencing in conjunction with underpasses is considered an effective way of ameliorating the impact of roads for species using the underpasses.

Australian Museum Business Services Consulting (June 2002) Fauna underpass monitoring Stage Two, Episode Five, Taree, NSW Roads and Traffic Authority

Location of study: Pacific Hwy, Taree, NSW

Habitat: Dominant communities are open dry sclerophyll forest, wet sclerophyll forest, swamp sclerophyll forest and mangroves. Dominant species are also given.

Target species: Small mammals, introduced carnivores, arboreal mammals, some reptiles

Mitigation structure/s: Three underpasses. One 47.8 m long and 2.8 m wide steel arch design with a small skylight. One single-span bridge, 10 m long, with a 3 m riverbank and 1 m wide concrete shelf, leaving approximately 4 m available for movement under the bridge (the 1 m wide concrete section was monitored). One 1.35 m diameter 62.2m long multiple-pipe drainage culvert (this culvert did not include fauna


exclusion fencing as it was intended for drainage rather than wildlife). Details of wildlife exclusion fencing are referenced in another report.

This report contains monitoring from Taree during Stage Two, Episode Five of a larger project spanning several reports. Monitoring for Episode Five occurred from 11 November 2001 to 16 December 2001. This report also included broader information on the study as a whole, which involved monitoring the use of culverts by a variety of species under a 23-km stretch of the Pacific Highway. The entire study took place over 18 months (July 2000 - December 2001).

Three underpasses were studied during Episode Five using a combination of sand trays and cameras linked to infra-red beam triggers (as described in previous reports). Vegetation descriptions were obtained from previous studies. The number of complete passages made by animals of different body sizes were compared using a Chi-squared analysis. Underpass attributes could not be analysed due to lack of replication.

The most common species recorded in the three culverts during Episode Five were Northern Brown Bandicoots (Isoodon macrourus), Black Rats (Rattus rattus), and antechinus (includes Yellow-footed, Antechinus flavipes, and Brown Antechinus, Antechinus stuartii) respectively. In total, thirteen native and six introduced species were recorded during the Episode, with the Koala the only threatened species. Species level identification was not possible for antechinus due to frequent flooding preventing camera installation. Two of the underpasses had less complete passages by larger species (e.g., koalas and wallabies), while more of these moved through the third underpass than smaller species. A single Blue-tongue Lizard was recorded killed on the road on the Taree bypass during Episode Five.

The study confirmed that a variety of species use the underpasses. Use varied over time. For example, more reptiles were recorded in summer, and more koalas, antechinus and bandicoots were recorded around their breeding seasons. The presence of stumps may have deterred some species. Surrounding suitable habitat is also likely to affect what species are using the structure. The largest underpass was used by the greatest variety of species and was the only structure used by larger species.

Hamer, A. & Organ, A. (July 2006) Craigieburn Bypass: monitoring of crossing structures for the growling grass frog


Litoria raniformis, Thomastown to Craigieburn, Victoria, Ecology Partners, Final report for Vicroads

Location of study: Hume Freeway, Craigieburn, Victoria

Habitat: Wetlands. Dominant emergent species include water ribbons (Triglochin procera), tall spike rush (Eleocharis sphacelata), common reed (Phragmites australis) and cumbungi (Typha sp.). Submerged species include Potamogeton sp. and Myriophyllum sp.

Target species: Growling Grass Frog, Litoria raniformis

Mitigation structure: Six frog underpasses and four combined culvert and underpass structures. Dimensions not given.

This report was a survey for the endangered Growling Grass Frog around culvert and underpass structures under the 17 km long Craigieburn Bypass. Surveys for frogs, and habitat assessment were conducted, and frog fences adjacent to crossings inspected. Surveys were conducted at 14 ponds at either end of 6 frog underpasses, 9 ponds at either end of 4 combined drainage/frog underpasses, 13 waterbodies/wetlands within 50 m of underpasses, and 5 creeks/drainage lines. Nocturnal spotlighting was used to detect frogs, for 7 nights over a 3 week period (24th November 2005 - 15th January 2006). Surrounding areas were not surveyed. Spotlighting took 5-20 minutes per site, varying with size and habitat complexity. Each site was surveyed a different number of times, with sites considered to have appropriate habitat surveyed more rigorously.

No Growling Grass Frogs were recorded from any of the underpasses or ponds at their entrances. Two males and a female were recorded at a wetland within 50 m of one underpass. Waterbodies and wetlands created to intercept road runoff were considered high habitat quality for the species, but suitability of the water quality for breeding and recruitment at these sites was unknown. Eight underpasses are thought to have ponds that are too small in size and depth (as they do not hold water permanently) or contain too much rock and too little aquatic vegetation.

The frogs recorded are believed to be isolated from adjacent sub-populations, therefore maintenance of underpasses is recommended. Suggestions made for improving crossing structures include rock removal, increasing pond size, planting submerged vegetation in created wetlands, and weed removal. Trees and shrubs should not be


planted in the vicinity of ponds, as shade reduces habitat suitability for the species. The maintenance of existing habitat is stressed for underpasses to be effective. The creation of additional ponds near underpass entrances is also suggested as an option for attracting frogs and increasing their movement. The importance of drift fencing being secured into the ground to prevent frogs entering the road is mentioned. Further surveys in the area are also recommended, extending to more activity periods, a wider survey area, and individual tagging to identify frog movements.

Jones, D., Appleby, R., Edgar, J. & Green, B. (October 2004) Compton Road wildlife movement monitoring program: preliminary results of Phase 1, A report for Environment and Parks, Brisbane City Council, Suburban Wildlife Research Group

Location of study: Compton Rd, Kuraby, Queensland

Habitat: Karawatha Forest Park, vegetation not described.

Target species: Birds, arboreal mammals, frogs and reptiles

Mitigation structure/s: n/a

A preliminary report on fauna surveys conducted prior to a road upgrade along 1.3 km of Compton Road. The road upgrade will include a number of mitigation structures to facilitate wildlife crossing. This report forms the beginning of a longer-term project aimed at collecting quantitative data on a variety of species on either side of the road before, during and after construction. It is suggested that mitigation structures will minimise road-kill, but these are not detailed in this report.

Monitoring was conducted over 4 months (February to 1st June 2004), running until the start of construction activities. Road-kill was assessed using twice-weekly foot-surveys along the both sides of the entire length of road. Bird surveys were conducted along 5 transects on either side of the road, within 500 m of the road. Fixed-width transects were used (100 x 15 m length) immediately followed by 5-minute point count surveys. A total of 16 bird surveys were conducted. Arboreal mammals were surveyed using nocturnal spotlighting transects, with six 5-6 hour long surveys conducted. Pit-fall trapping for frogs and reptiles is planned for future phases of the project.

Eighteen species were found killed on the road. The majority of


mortality was comprised of Cane Toads (over half of individuals), but 14 native species were also present. The majority of these were frogs and reptiles, such as the Striped Marsh Frog (Limnodynastes peronii). Bird surveys detected 72 species, with similar species present on either side of the road. At least 60 individuals of 8 species of arboreal mammals were observed, most abundant being the Little Red Flying Fox, Common Brushtail Possum, and Common Ringtail Possum.

These data are considered base-line information for future phases of the project, and no statistical analyses or recommendations are included at this stage. It is suggested that road-kill levels are relatively high, and the surveys identified a high diversity of arboreal mammals including locally significant species (Koala and Greater Glider).

The results of subsequent monitoring of the mitigation structures is detailed in Bond and Jones (2008) and is summarised there.

Ecologia Environmental Consultants (June 1995) Kwinana Freeway wildlife underpass study, Fauna monitoring program, Main Roads Western Australia

Location of study: Kwinana Freeway, Perth, WA

Habitat: Bassendean vegetation system. Banksia woodland is less widespread in the southern region. On more moist soil it merges into Jarrah- Marri woodland, varying from open trees with at thick Banksia understorey to a well-formed woodland with scattered smaller Banksia and Casuarina. Freshwater swamps between dunes occupy a large proportion of the system. The area has a dry Mediterranean climate.

Target species: Southern Brown Bandicoot (Isoodon obesulus), other native and exotic mammals, reptiles

Mitigation structure/s: Six 600 mm diameter concrete pipe underpasses, of varying length.

This report was a survey of animals in and around underpasses along the Kwinana Freeway, WA, between Forrest and Thomas Roads. Six underpasses were located along this stretch of road, intended to restore connectivity for wildlife. Exclusion fencing was present along the road, though details of this are not given.

Surveys were conducted during December 1994 and March/April 1995. During these survey periods, activity in and around underpasses was


detected from diggings, scats and hairs. Sand trays and soot pads were installed to detect tracks. An infra-red motion sensor was also used at underpass entrances, connected to a mechanical counter and camera. Surveys were also conducted in the area surrounding the underpasses using trapping, and habitat analyses conducted to determine preferred habitat, and expected underpass use. Road mortality was also recorded.

Sand trays were placed at all access points, and soot pads at strategic points along the length of underpasses. These were monitored daily, and direction of movement recorded. Trapping of individually marked individuals was conducted on either side of underpasses. A grid of 20 baited Elliot and cage traps was placed in two lines of 10 traps, 25 m apart, running parallel to the freeway. Trapping was conducted over three consecutive nights at each underpass, then rotated. This occurred during the two survey periods (summer and autumn). The sensors and cameras were rotated between underpasses during the study period.

Of the species detected in the area only four were recorded moving through underpasses: Bobtail (Tiliqua rugosa), Gould’s Monitor (Varanus gouldii), Feral Cat (Felis catus) and Southern Brown Bandicoot (Isoodon obesulus). Most tracks found at underpass entrances were restricted to the entrance, with only 18 per cent continuous through the tunnel. Other species trapped nearby included House Mouse (Mus musculus), Black Rat (Rattus rattus) and Skink (Ctenotus fallens). Tracks, road-kill and opportunistic observations also revealed the presence of Rabbits (Oryctolagus cuniculus), Fox (Vulpes vulpes), Dugite (Pseudonaja affinas), Dog (Canis familiaris), Brushtail possum (Tricosurus vulpecula) and Australian Raven (Corvus coronoides). It is suggested that there is a high use of underpasses by feral animals. Use by native fauna appears dominated by reptiles in summer, and bandicoots in autumn. The exclusion fencing present was ineffective at keeping some species off the road, with a monitor observed attempting to climb the fence.

The low-incidence of underpass use observed meant than only twelve data sets were available for analysis. Univariate, simple linear regressions were conducted on a variety of indices of underpass use/performance compared to biophysical variables. A multiple regression analysis was carried out on variables found to have a significant relationship with underpass performance. A stepwise regression was also carried out on these variables. Significant relationships with underpass performance were found with distance to


vegetation, leaf litter distribution, shrub cover (0-0.5 m tall), and shrub cover (1-2 m tall).

The rehabilitation of areas surrounding underpasses is recommended to maximise the presence of fauna in the immediate area, and therefore likelihood of underpass use. This includes rehabilitating vegetation and ensuring it is present at underpass entrances, and returning vegetation debris, logs and rocks to disturbed areas. Ongoing monitoring to determine underpass effectiveness is also recommended. It is suggested that underpasses are currently functioning at an ‘acceptable level’ for Southern Brown Bandicoot, with the species accounting for nearly half of all fauna visiting tunnel entrances, with a small proportion passing through. It is predicted that this will increase as individuals become accustomed to underpasses. It is possible that feral animals are deterring bandicoots from some underpasses.

Harris, I. (2007). Use of wildlife underpasses by quenda (Isoodon obesulus fusciventer) at Roe Highway Stage 7. Honours Thesis, Faculty of Natural and Agricultural Sciences, The University of Western Australia

Location of study: Roe Highway Stage 7, Perth, WA

Habitat: relatively undisturbed Banksia dominated woodland at the transition between the Spearwood and Bassendean dune systems

Target species: Southern Brown Bandicoot (Isoodon obesulus)

Mitigation structure/s: two purpose built box culverts (1.2 m x 1.2 m x 35 m long and 1.2 m x .45 m x 20 m long) and one modified utility culvert (1.2 m x 1.2 m x 45 m long).

Investigated the use of three box-cell culverts under a new section of the Roe Highway within suburban Perth. The surrounding banksia woodland contained populations of Southern Brown Bandicoots (also known locally as quenda) and mitigation was specifically targeted to them. Monitoring (daily) during the first 12 months (August 2005 – August 2006) was by identifying tracks in sand pads. The floor of the two purpose-built culverts was completely covered with fine sand allowing for complete passes to be detected. The third culvert had a strip of sand at each end. From August 2006 – August 2007, the use of the most frequently-used culvert was monitored using a scanner that detected animals that had been fitted with microchips during trapping.


Various technical difficulties resulted in some lost data.

Bandicoots were trapped around the road in both years of the study. A total of 278 complete passes were detected using the sand tracking pads from 2005 – 2006, with approximately two-thirds originating from one culvert. Seasonal differences in rate of use in the first 12 months were evident. Eight individuals (4 male and 4 female) were detected using the underpass monitored with the microchip scanner. This represents 28 per cent of the tagged population in 2006. At least 6 of the 8 individuals used the underpass regularly and repeatedly, suggesting it was used for daily movements by these microchipped animals. Foxes were also reported using the three culverts, potentially resulting in increased rates of predation on the bandicoots. While no direct evidence of predation was obtained, the sudden cessation of use of the culverts by multiple individual bandicoots suggests that predation may be an issue.

Australian Museum Business Services Consulting (August 1997) Fauna usage of three underpasses beneath the F3 Freeway between Sydney and Newcastle, Roads and Traffic Authority

Location of study: F3 Freeway between Sydney and Newcastle, NSW

Habitat: The area includes the overlap of two zoogeographic zones. The Torresian (wet tropic) zone supports tall eucalypt forest on rich, loamy soils, with stunted forest and heath on sandy or stony soils. The Bassian (temperate) zone is dominated by eucalypt forests with low-diversity rainforest in moister areas. The area as a whole has low elevation and generally broken topography, with a mild, coastal climate.

Target species: Small to medium-sized mammals, reptiles

Mitigation structure/s: Three underpasses. One 10 m diameter tunnel. A site with three box culverts 3 m wide x 1.5 m high. A four-cell pipe 1.5 m diameter. Structure lengths not given.

Investigated the use of three underpasses passing under the F3 Freeway between Sydney and Newcastle. Three underpasses were selected for monitoring, based on ease of access, and the presence of appropriate habitat nearby for target species. The underpass in Brisbane Water National park was purpose-built for fauna, whereas the culvert near Warnervale and pipe in Awaba State Forest were primarily for drainage. One underpass had exclusion fencing aimed at keeping


animals off the road and funnelling them towards the underpass.

The underpasses were monitored over 9 months (August 1996 - June 1997). A camera connected to an infra-red sensor was set up at each underpass to record species use. Its placement was aimed at detecting complete passages only. 1-2 m wide and 5-10 cm deep sand beds were also set up approximately 2 m inside both entrances of the underpasses, extending their entire width. During site visits any other signs of animals were noted, such as scats, hair or tracks elsewhere in the underpass. The road was checked for road-kill 100 m either side of the underpass at each site visit, with other road-kill data coming from liaison with other organisations.

The expected presence of fauna in the area of underpasses was determined using information from databases and literature searches, focusing on small to medium-sized mammals, both ground-dwelling and arboreal. There was also a specific focus on information relating to endangered species. This information was complimented with road-kill data, opportunistic observations, and limited surveys using trapping and spotlighting. Animals too small to be detected by the infra-red sensor were not included. In areas surrounding underpasses trapping was conducted over four nights in August 1996, using a combination of Elliot and cage traps. Traps were located along transects perpendicular to the freeway, approximately 5 m apart. Arboreal species were surveyed using trapping and spotlighting during February/March 1997.

Photographs recorded 17 native species using the underpasses. Thirteen were recorded at Brisbane Water National Park, with Common Wombat (Vombatus ursinus), Swamp Wallaby (Wallabia bicolor), Rats (Rattus spp.) and Long-nosed Bandicoots (Parameles nasuta) the most common. Five native species were found using the culvert at Warnervale, and seven in the drainage pipe at Awaba State Forest. The species recorded represent a subset of the fauna present in the area, though use by a large range of fauna was shown to use the underpasses. Some species were recorded more during summer, particularly reptiles. The only threatened species detected using the underpasses was the Spotted-tail Quoll (Dasyurus maculatus), recorded once only. It is not thought that predators are using underpasses as prey traps, as there were infrequent records of their passage and no evidence of carcasses, but this study did not provide conclusive proof that this is not occurring.


140 road-kills were recorded during the study, consisting of individuals of 43 natives and 43 introduced predators. Most frequent was the Australian Magpie. Exclusion fencing was not preventing freeway access, though established pathways suggested that it was funnelling some individuals towards the underpass. Signs of several species were present between the fencing and freeway. This was possibly due to breaches in the fence, or the fact that fencing did not enclose the entire road easement.

The report gave preliminary recommendations to incorporate a maintenance plan for exclusion fencing as part of road maintenance, use of native grasses in landscape planting, ensuring that underpasses are of a size allowing the skyline to be seen from the other end, designs that allow dry passage, and taking into account adjoining habitat and land use when locating underpasses. A literature review was also used to make more detailed recommendations.

Bax, D. (February 2006) Karuah Bypass: fauna crossing report, Thiess Pty Ltd for Roads and Traffic Authority.

Location of study: Pacific Highway, Karuah Bypass, north of Sydney, NSW

Habitat: Not described

Target species: Gliders, and other arboreal mammals

Mitigation structure/s: One of five canopy bridges, elevated above the road and suspended between two poles on either side of the road. Rope is woven into a 300 mm wide x 200 mm high rectangular tube. Stainless steel frames keep the tube in shape. Poles were placed as close as possible to existing trees, and ropes were draped to adjacent vegetation. Further technical details are provided.

Investigated the effectiveness of canopy bridges across the Karuah Bypass, which was constructed between January 2002 and December 2004. The five crossings designed specifically for gliders and other arboreal mammals were installed along the 9.8 km stretch of road. Only one overpass was monitored and the results presented in this report.

The eastern side of one overpass was monitored using a motion-detecting camera system. The camera was in place for 8 months (April 2005 to December 2005). The sensor was set up using an arrangement of mirrors allowing movement at both the top of the rope tunnel and


inside it to be detected. An infra-red flash was used to avoid startling fauna.

During the monitoring period, animals were photographed 50 times. Forty-six of these were the Common Brushtail Possum, and four Squirrel Gliders. This confirmed that possums and gliders will cross aerial rope bridges up to 70 m in length with the design used. Placing the crossing ends close to vegetation is thought to be beneficial. The design may be simplified to a ladder instead of a tube to save costs, as most animals crossed on top rather than inside the rope-ladder. Aerial ropeways are advocated as better than median-strip trees because they also allows non-gliding fauna to cross, assist monitoring, and reduce the likelihood of animals going onto the road from the median strip. Not requiring trees in the median strip may also benefit road safety, and reduce the overall footprint of the road,

Some recommendations for improving the monitoring technique used are also discussed, in order to minimise false triggering and enable complete or partial crossings to be distinguished.

Fitzgerald, M. (March 2003) Results of fauna monitoring in designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: February-March 2003

Location of study: Pacific Highway, Yelgun to Chinderah, NSW

Habitat: Agricultural and pastoral areas dominate in the north, further details are not given.

Target species: Mammals, birds, reptiles, amphibians.

Mitigation structure/s: Thirteen crossing structures. Includes seven bridges, two arches, one cut-and-cover tunnel and three reinforced concrete box culverts. Dimensions not given.

Monitoring of fauna crossings through twelve structures under the Yelgun to Chinderah Pacific Highway Upgrade and one under the old Pacific Highway.

Sand traps were set up and monitored 2-3 times per week between 10 February and 31 March 2003. The configuration of sand traps varied for each crossing, most contained 2-3 sand traps with a few continuous sand traps. Where possible, traps ran the full width of the underpass, though this was not possible where there were watercourses or steep


areas. Scats and road-kill were also recorded as part of monitoring. Information about fauna likely to be present in the area was obtained from previous surveys.

Twenty-two fauna groups were identified using the underpasses, from 1248 recognisable tracks. This included fourteen common natives and eight introduced species. The most common species recorded was the Cane Toad (Bufo marinus), which was present at 11 of the crossings and considered ‘conspicuously abundant’. The most common traverses of underpasses that could be distinguished were by Swamp Wallabies (Wallabia bicolor), followed by dogs. No threatened species could be identified, though bird tracks may have included threatened species e.g., Black Bittern and Bush Hen. Road-kill recorded included Cane Toads, a variety of frog species, reptiles, Swamp Wallaby, echidna, hare, fox and birds.

It is suggested that fauna recorded varies with quality, extent and interconnectivity of adjacent habitats, and ability to access crossings. Mice and Cane Toads were more abundant in more northern crossings, possibly due to higher levels of disturbance in the surrounding landscape at these locations. Avoidance between fauna groups may also be occurring, e.g., alternating dominance between water dragons and water skinks. Other influences considered likely are state of plantings, episodic flooding and surrounding landuse. Further influences are hypothesised for future monitoring, including time since construction, seasonal variations and effects of established plantings.

Fitzgerald, M. (January 2004) Draft report, results of sandtrap monitoring in fourteen designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: Sample 2 October 2003-January 2004


Habitat: Not described.

Target species: Mammals, birds, reptiles, amphibians

Mitigation structure/s: Fourteen underpasses, consisting of two arches, three reinforced concrete box culverts, two cut and cover tunnels and seven bridges. Dimensions not given.

Continued monitoring of fauna using underpasses along the Yelgun to Chinderah Pacific Highway Upgrade. This report outlines Sample Two


monitoring, conducted from October 2003 to January 2004. Sand traps were installed in the same way, and with the same monitoring frequency previously used (see description in Fitzgerald 2003 report).

During Sample 2 twenty-six fauna groups were identified from 2388 recognisable tracks. Eight of these were of introduced species, and two groups were likely to contain both native and exotic species (rats and small mammals). The most frequently recorded complete crossings were by Swamp Wallabies. The fauna recorded were essentially the same species and groups identified during Episode One. Tracks of one threatened species were found: Spotted-tail Quoll. There was a positive correlation between occurrence of fox and cat tracks, suggesting they may favour similar area, habitat, prey resource, or crossing attributes.

Fitzgerald, M. (April 2005) Final report: results of sandtrap monitoring in eight designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: Sample 3 February-April 2005, for AbiRoad Maintenance Pty Ltd.


Habitat: Not described.

Target species: Mammals, birds, reptiles, amphibians

Mitigation structure/s: Two cut and cover tunnels, two bridges, one arch and three reinforced concrete box culverts. Dimensions not given.

Continued monitoring of fauna using underpasses along the Yelgun to Chinderah Pacific Highway Upgrade. This report outlines Sample Three monitoring, conducted from 23 February to 8 April 2005. Sand traps were installed in eight crossings, selected based on areas of habitat quality. They were set up in the same way, and with the same monitoring frequency previously used (see description in Fitzgerald 2003 report). Comparisons are made with previous monitoring and conclusions drawn from results from all three survey periods.

During Sample Three, twenty-three fauna groups were recorded from 1190 recognisable tracks. 111 complete crossings by 18 fauna categories were recorded, though this is thought to be an underestimate, particularly where wind interfered with sand traps. The most frequent crossings were by Swamp Wallabies, dogs, snakes and rats. Native fauna predominated over introduced species. Introduced carnivores were present at all eight crossings. Cane Toads were also


recorded at all underpasses. A koala was recorded, confirming another vulnerable species was present. Tracks and scats present away from the sand traps assisted in species identification but did not reveal the presence of any additional species. Seven road-kills were recorded, consisting of several bird species, and a Mountain Brushtail Possum (Trichosurus caninus).

Similar fauna was found in Samples One and Two. Species using underpasses were predominately a subset of common vertebrate species of mixed agrarian, regenerating forest and native forest landscapes.

Factors identified as likely to be affecting the proportion of traverses include state of adjacent vegetation, types of fauna present in this habitat, their abundance and readiness to enter culverts. Detectability was also greater in sheltered areas due to better preserved tracks. Time for vegetation establishment, assisted by cattle exclusion and tree planting, are also likely to have contributed to fauna usage. Over time crossings may become more familiar and so more likely to assist connectivity. The barrier effect of roads depends on the health of the surrounding landscape and associated faunal abundance.

Hayes, I.F. (2006) Effectiveness of fauna road-kill mitigation structures in north-eastern New South Wales. Unpublished Third Year Undergraduate Report. School of Environmental Science and Management, Southern Cross University, Lismore.

Location of study: Pacific Highway, Yelgun to Cudgera Creek and Woodburn to Ferry Park Rest Area, north-eastern NSW

Habitat: Area of overlap between temperate and tropical zones. 1580 mm mean annual rainfall. Sediments mainly of sandstone, conglomerate, siltstone and coals. The area includes rainforest and wet and dry sclerophyll forests. Dry hardwood forests occur on poorer soils. Dominant species are given. Much land has also been cleared for agriculture.

Target species: Predominantly mammals, some reptiles and amphibians

Mitigation structure/s: Monitored two underpasses and two overpasses. Included a dedicated fauna overpass (cut and cover tunnel) and dedicated fauna underpass (3 x 3 m concrete box culvert). Further


dimensions not given.

Investigation of road-kill along two sections of the Pacific Highway. Also reviews the ecological impact of roads, use of roadside corridors by wildlife, details of potential mitigation structures and the history of road-kill studies in Australia. The Yelgun to Cudgera Creek section contains 47 structures that may mitigate road-kill, including one dedicated fauna overpass. The Woodburn to Ferry Park Rest Area section does not contain any dedicated fauna crossing structures, but drainage culverts may be used by smaller species. Exclusion fencing was present along the Yelgun to Cudgera section, with a total of 6,150 m of fencing along the northbound carriageway kerbside, and 5,700 southbound. This included sections 2.1 m and 1.4 m in height.

Two underpasses and two overpasses were monitored between Yelgun and Culdgera Creek. Sand plots 1 m wide and 20 cm deep were installed the full width (3 m) of the eastern entrance of structures and checked up to three times per week for six weeks (August-September 2006). The direction of movement was recorded where possible. The plots were smoothed over after each visit.

Road-kill was also monitored over 7 weeks (August-September 2006) along both sections of the highway. Monitoring occurred once per week along the Woodburn to Ferry Park section, and up to three times per week along the Yelgun to Cudgera Creek. Extra data on road-kill were incorporated from Abigroup Contractors Pty Ltd. Data on the number and location of fauna warning signs were also recorded.

At least eight non-flying mammals, four reptiles and one amphibian species were recorded using the structures. More large animals (e.g., macropods, canids, large lizards) were found using overpasses than underpasses. Bandicoots, rodents and amphibians were more commonly detected in underpasses, though weather may have obscured some tracks on overpasses. Possible factors identified as influencing the use of structures are vegetative cover at the entrance, nearby habitat connectivity, structure length, weeds impeding movement and the presence of predators.

78 road-kills of 21 identified species were recorded in total, most of which were mammals. Mortality was higher on Woodburn to Ferry Park section than Yelgun to Cudgera Creek (15 and 12 species respectively). On both sections, the most common species found killed on the road was the Northern Brown Bandicoot (Isoodon macrouris). Six road-killed


mammals were found inside exclusion fencing. No fauna warning signs were present between Yelgun and Cudgera Creek. Those installed on the Woodburn to Ferry Park section are considered adequate, though their effectiveness is unknown. It is suggested that exclusion fencelines were weedy and arboreal mammals were able to climb the fences, reducing the effectiveness of these between Yelgun and Cudgera Creek. Regular maintenance is recommended.

Suggestions for future research are made, including more comprehensive surveys, and investigation of population dynamics to determine if mortality rates are affecting the viability of local populations.

Scotts D. J. & Seebeck J. (1989) Ecology of Potorous longipes (Marsupialia: Potoroidae); and preliminary recommendations for management of its habitat in Victoria. Arthur Rylah Institute for Environmental Research Technical Report Series No. 62. Department of Conservation, Forests and Lands, Heidelberg, Victoria. (information detailed in Appendix 2 of report)

Location of study: Princes Highway near Bellbird, East Gippsland, Victoria.

Habitat: Eucalyptus forest

Target species: Long-footed Potoroo

Mitigation structure/s: Three corrugated metal round culverts (2 x 1050 mm diameter and 1 x 900 mm diameter), each positioned where road design required structural filling. Gravel and dirt was placed within each culvert and one culvert was monitored using remotely triggered camera from January to April 1987. Subsequent monitoring of all three culverts using sand tracking pads was sporadic and have not been formally reported (Tony Mitchell, DSE Orbost, pers comm.).

No Long-footed Potoroos were observed using the culvert either using photography or by detection of tracks. Wombats, Bush Rat, Dusky Antechinus, Common Ringtail Possum and unidentified bird and bat were detected within the culverts by footprints and camera by Scotts and Seebeck as well as subsequent monitoring by Tony Mitchell. Concern was raised about the suitability of adjacent habitat to support Long-footed Potoroos.


van der Ree R., Soanes K., Cesarini S., McCall S., Taylor A., Sunnucks P., Harper M., Gulle, N. (2008). Rope bridge story revisited. Episode 30 of The University of Melbourne 'Visions' Video podcasts. www.visions.unimelb.edu.au accessed 12 August 2008.

Location of study: Hume Highway, between Seymour and Benalla, NE Victoria.

Habitat: Eucalyptus woodland

Target species: Squirrel Glider, Common Ringtail Possum, Common Brushtail Possum, Yellow-footed Antechinus

Mitigation structure/s: Two canopy rope bridges (70 – 80 m long) and three glider poles positioned within centre median.

This video podcast is a brief summary of the findings of research that aimed to quantify the barrier effect of major roads for a range of possums, gliders and skinks (skinks not reported on) and the use and effectiveness of mitigation measures for possums and gliders. The rope bridges were installed in mid 2007, and this video summarises the first 12 months of monitoring using remotely-triggered infra-red cameras at each end of both rope bridges. After approximately 12 months of monitoring, Common Ringtail Possums were using the structures regularly, with 50 complete passes recorded. Common Brushtail Possums and Squirrel Gliders have been detected on the bridges seven and four times each, respectively, but complete traverses have not yet been detected. There has been no monitoring of the glider poles yet. Extensive surveys and pre-mitigation research was carried out but was not reported in the podcast. Monitoring of use and evaluation of effectiveness is ongoing.

8.3 Australian ‘Best Practice’ manuals and Guidelines

Department of Environment and Conservation (2006) Avoiding and Off Setting Biodiversity Loss: Case Studies, Department of Environment and Conservation: NSW

Location of Study: Wallarah Peninsula, Karuah Bypass and Federal Hwy Canberra

Habitat: 600 ha of undisturbed Coast and Lake Bushland at Wallarah Peninsula, Natural Temperate Grasslands along the Federal Hwy and 47


ha of natural bushland within the Karuah Nature Reserve, Karuah.

Target Species: Striped Legless Lizards, and a number of threatened species in general.

Mitigation Structures/ Measures: For the Karuah Bypass the following structures were implemented:

Several dedicated fauna underpasses as well as combined drainage and fauna underpasses.

Dry passage access for fauna under major bridge crossings.

Floppy top fauna exclusion fencing along the boundary of Karuah Nature Reserve.

Retaining native vegetation in the median strip to allow for glider access

Installing experimental rope ladder/tunnel ‘glider crossings’ at some points

Replanting disturbed areas with native species.

Installing fencing round threatened flora species to protect against accidental damage during construction.

For the two other case studies the mitigation measures involved compensatory habitat in alternative locations to the original habitat and establishing environmentally sensitive designs to allow roads, infrastructure and habitat to co-exist.

Major Linear Infrastructure: Federal Hwy, Pacific Hwy, and Residential roads and properties in the Wallarah Peninsula.

This report is a practical guide written by the Department of Environment and Conservation (DEC) for developers, governments and other infrastructure related groups, to assist with dealing with potential environmental impacts. In this report DEC has chosen three representative case studies: (Wallarah Peninsula, Federal Hwy and Karuah Bypass) to demonstrate the potentials for developers and landholders to either completely avoid or to compensate/ mitigate biodiversity impacts when looking to develop infrastructure in sensitive environmental areas. From the review of these case studies the DEC has proposed to develop a biodiversity offset and Bank scheme to fast track the process of development.

Each case study was used to demonstrate the different lessons that can be learnt when dealing with these problematic environmental issues.


The outcomes of each study are as follows:

The Wallarah Peninsula case study demonstrates a case where impacts could be avoided and minimised without resorting to biodiversity offsets. The developer recognised that biodiversity was an asset to the area and established an environmentally sensitive residential development.

In the case of the Karuah Bypass, the Roads and Traffic Authority (RTA) acknowledged that it could not avoid, minimise and mitigate all the impacts on biodiversity on-site. The RTA provided 89 hectares of compensatory habitat to offset the loss of habitat and implemented the above mentioned mitigation structures (see also the summary of the report by Bax 2006).

In the Federal Highway upgrade project, the RTA purchased a property with a known population of Striped Legless Lizards and protected this land from development, to offset impacts on a smaller population at another location.

Furthermore, whilst these case studies demonstrate good outcomes, negotiating biodiversity offsets the DEC recognised that determining solutions to such problems on a case-by-case basis can be resource intensive and slow, and that there is potential for inconsistency in the process and outcome.

Based on the recognition that dealing with such problems on a case by case basis is time consuming, DEC introduced the idea of developing a biodiversity offsets and banking scheme to fast-track the process within this report. DEC proposed that the scheme would do the following:

Address the impacts of development on biodiversity values;

Recognise the market values of biodiversity;

Create new opportunities for conservation management on privately-owned land to complement the State’s national parks and other protected areas; and

Provide transparent, consistent assessment procedures and defined ecological principles for offsetting.

This publication broadens the scope and solutions to environmental impacts of major infrastructure and provides adequate detail on each case study. However, it is deficient in the detail of whether these compensatory/mitigating measures solved or ameliorated the potential


direct or indirect impacts on biodiversity that the projects may have had. Additionally this document lacked distinctive conclusions and yet still proposed to create a scheme to fast track the system of dealing with environmental impacts, so that individual cases could be processed without being looked at independently.

Goosem, M W (2005) Wildlife surveillance assessment Compton Road Upgrade 2005: Review of contemporary remote and direct surveillance options for monitoring. Report to the Brisbane City Council. Cooperative Research Centre for Tropical Rainforest Ecology and Management. Rainforest CRC, Cairns, 69 pages.

Location of Study: International review of literature and targeted questionnaires to practitioners of methods used to monitor the use of wildlife crossing structures.

Habitat: terrestrial habitats

Target Species: Typically terrestrial vertebrates.

Mitigation Structures: All types of overpasses and underpasses.

Major linear Infrastructure: Typically roads

This review was undertaken for the Brisbane City Council to guide them in their evaluation of the use and effectiveness of the mitigation structures along the upgraded Compton Rd. The authors examined the international literature and contacted all authors to identify the range of survey techniques and evaluate their efficacy. The report details the different types of equipment, an indication of costs (2005 data) and some of the limitations of the various techniques. Recommendations specific to the mitigation structures used at Compton Road are given.

The general conclusions of the review were:

� Confounding variables that influence rates of use of structures by wildlife need to be considered. These variables include presence / activity of humans, vegetation condition, structure type/dimension.

� Abundance of animals adjacent to the structures will also influence rate of use, and simultaneous monitoring of populations is necessary

� Seasonal effects and time since construction may influence rate of detection and should be considered

� Monitoring should rely on a combination of methods to maximise the number of species detected and provide a back-up in the case of failure of electronic equipment


Main Roads Western Australia, (June 2002) Design of Fauna Underpasses, pp8.

Location of Study: General summary not specific to any locations.

Habitat: Terrestrial habitat impacted by major linear infrastructure.

Targeted Species: Not specific but mainly larger terrestrial species.

Mitigation Measures: Underpasses of varying types.

This publication is the first of two by Western Australian Department of Main Roads and provides general information on the design of fauna underpasses. It mentions, in broad terms, the impacts of habitat fragmentation on fauna populations.

It suggests that fauna underpasses should be designed to cater primarily for the fauna species at risk of habitat fragmentation. It states that the dimensions of underpasses should be designed relative to the size of the animals that will use them. It advises that information on the species could be obtained by conducting fauna surveys and suggests that the Department of Conservation and Land Management (CALM) are able to provide advice on fauna surveys in order to assess the species, population size and preferred treatment options when existing habitats are impacted by road construction works.

The publication mentions fauna underpasses and discusses the role and reaction of certain animals to fences.

The publication discusses and makes useful suggestions of means to encourage fauna to use underpasses

The publication is a very general overview.

Main Roads Western Australia, (June 2002) Design of Fencing/Walls, pp10.

Location of Study: General summary of types of fencing and their uses

Habitat: N/A

Targeted Species: It is not specifically related to fencing for fauna

Mitigation Measures: Fences

This publication is the second of two by Western Australian Department


of Main Roads. It provides a listing of the types of fences for all purposes. It gives a useful overview of the uses of each type and gives specific drawing codes for each fence type. It gives advice on the siting and erection of fences. It gives indications of the factors to be considered for different types of fence.

The publication mentions a range of other issues relating to fences other than for fauna and wildlife, such as drainage, OH&S, general protection, high risk areas, visual and screening, and being adjacent to high voltage power lines.

It discusses the practicality and economics to fencing highways and main roads in pastoral areas, mentions factors, eg., road reserves and stock grids. The publication also mentions and lists a number of ‘standard fence’ drawings for use in pastoral settings. The publication also discusses other issues related to pastoral fencing, such as water severance and fencing underpasses.

The publication also mentions other issues not related to fauna such noise barriers and visual screening.

The publication is a useful practical guide to the selection of appropriate fences.

Qld Main Roads (2002) Fauna Sensitive Road Design V1: Past and Existing Practices, Qld Department of Main Roads, Planning, Design and Environmental Division.

Location of Study: Bruce Highway Kinhill, Sunshine Coast and F3 Freeway, Newcastle to Sydney

Habitat: Habitats not described, except terrestrial habitat that runs parallel to the two major roads.

Targeted Species: All terrestrial species. Field data based mostly on small- medium mammals

Mitigation Measures: Culverts, underpasses and bridges/ arches of varying size.

This publication is volume 1 in a two part series by Queensland Department of Main Roads (the second volume is yet to be published). This first volume provides background information pertaining to the effect that roads have on individuals, populations and the habitat of fauna species, by utilizing past and present examples. This review


discussed the following theoretical issues:

� Habitat fragmentation

� Island Biogeography

� Genetic Isolation, and

� Aspects of animal behaviour.

It then discussed the direct and indirect impacts of roads on fauna and identified practices currently used during road construction to help facilitate fauna movement. Additionally this publication reviews several scientific field results compiled from recent investigations of fauna use of underpasses and culverts.

The field data reviewed for this report appears to be limited due to small sample size, lack of controls and the generalist context of the results. This issue is acknowledged within the document when it stipulates that this review ‘has also established the need to undertake additional field surveys in a targeted manner and sound experimental design’.

QLD Main Roads, (1998) Roads in the Wet tropics, Planning, Design, Construction, Maintenance and Operation Best Practice Manual.

Location: Tropical northern QLD areas that receive greater than 1200mm rainfall per year

Target Species: Various

Mitigation structure: Closed canopies, pipe culverts

This is a ‘best practice manual’ which includes several guiding principles for the planning, design, construction, maintenance and operation of roads in tropical north QLD. The manual mentions several ways to reduce impacts on fauna that are caused by the construction and operation of roads.

The best practice manual provides guidelines which are presumably used by road designers and road maintenance personnel. The guidelines provided in the document that relate to mitigation measures for dealing with habitat fragmentation are summarised below;

Minimise habitat fragmentation and disturbance by facilitating fauna movement and, in closed canopy forests, promoting canopy closure


(p1).

To identify particular design elements which mitigate environmental impacts (e.g. fauna culverts etc.) (p3).

(provide) maintenance of drainage structures (particularly where they function as fauna culverts) (p3).

Vegetation around the inlet and outlets of the culvert should remain in a natural state to attract fauna (p5)

Large obstructions such as logs, branches, severe erosion and ponded water should be removed or positioned in such a way that allows fauna movement through the culvert whilst maintaining cover for fauna movements, (p5).

…ensure that relevant fauna management requirements are incorporated into (environmental) management plans.

While the publication provides the guidelines described above, very little discussion is provided on the design and construction of mitigation measures that are useful in mitigating habitat fragmentation. In particular, very little information is provided that would be useful to road designers using this document to mitigate habitat fragmentation.

While the publication is lacking in detail concerned with design and construction of habitat fragmentation mitigation measures, some useful information is provided on maintenance of fragmentation mitigation measures already in place.

Fairfull, S. and Witheridge, G. (2003) Why do Fish Need to Cross the Road? Fish Passage Requirements for Waterway Crossings. NSW Fisheries, Cronulla, 16 pp.

Location of Study: Australia wide- major waterways (e.g. Murray-Darling River), some research on four northern NSW rivers.

Habitat: Aquatic habitats across Australia.

Target Species: Freshwater Fish – with some focus on long distance migratory fish including Mary River Cod (30km), Silver Perch (570km), Murray Cod (1,000 km) and the Golden Perch which has been recorded swimming a staggering 2,300km.

Mitigation Structures: Bridge, arch structure or tunnel and culverts of varying size.


Major Linear Infrastructure: Roads

Basic technical report by NSW Fisheries providing guidelines for designing waterway crossings that allow continued fish passage across potential barriers such as roads, weirs and locks. The report is broken down into easy to follow sub-sections demonstrating why certain infrastructure (e.g. roads, weirs, fords) create a barrier effect to migrating fish and how changes in design/position of waterway crossings and mitigation structures (culverts, tunnels, bridges, arches etc) will better assist fish passage through these areas.

This report suggests an abundance of design considerations for ‘Fish Friendly’ water crossings. A summary of the publication conclusions are as follows:

Bridges and Arches: Avoid locating bridge piers or foundations within the main waterway channel, to allow continued flow of water and fish passage.

Culverts:

Minimise changes to the channel's natural flow, width, roughness and base-flow water depth through the culvert's wet cells. Wet cells should have a minimum water depth of 0.2-0.5 metres to encourage fish passage.

In all cases, the culvert should be designed to maximise the geometric similarities of the natural channel profile from the bed of the culvert up to a flow depth of 0.5 metres (‘Low Flow Design’).

Maximise light penetration within the wet cells by maximising the height or diameter of the cells, and possibly by introducing skylights or grated stormwater inlets into the median strip of divided roads. These skylights are only required within the nominated wet cells.

Fords: Where practical, the deck of the ford should follow the natural bed elevation.

Causeways: Minimise the use of causeways where fish passage is required.

During Construction: All reasonable and practicable measures should be taken to prevent or minimise environmental harm during the construction phase, including minimising restrictions of fish passage; minimising the release of sediment into the stream; minimising damage to, or the removal of, bank vegetation, particularly vegetation that


shades the low-flow channel.

Monitoring: Monitoring is highly recommended during both pre and post construction to ensure that the new waterway crossing design is successful in achieving the desired flow velocities and fish passage outcomes. The monitoring program should be designed in consultation with fisheries scientists and the local fisheries department/authority.

This publication is based on referenced scientific research. This publication is useful for the target audience.

Robertson P. (2002) Discussion Paper. Design requirement for structures to ameliorate the potential effects on frog movements of construction and operation of the proposed Craigieburn Bypass Freeway. Wildlife Profiles, Heidelberg, Victoria.

This report summarises the issues relating to the design of structures to facilitate the movement of the Growling Grass Frog across the Hume Freeway Bypass of Craigieburn. It summarises the knowledge at the time of the distribution, habitat requirements and ecology of the species at the site of the bypass. It then provides advice on a range of options to minimise the impacts of the highway on the species and facilitate connectivity, including the use and design of culverts, tunnels, drift fences, and monitoring and evaluation options.

8.4 International refereed journal publications

Dodd Jr, C.K., Barichivich, W.J. & Smith, L.L. (2004) Effectiveness of a barrier wall and culverts in reducing wildlife mortality on a heavily traveled highway in Florida, Biological Conservation, 118: 619-631.

Location of study: Paynes Prairie State Preserve, Florida, USA

Habitat: highland freshwater marsh (18.3 m above mean sea level), depending on rainfall and drainage, it may be a dry prairie, marsh or shallow lake

Target species: amphibians, reptiles, mammals, fish

Mitigation structure/s: Barrier wall-culvert system. Wall: 1.1 m high with a 15.2 cm overhanging lip. Eight concrete culverts in total: two partially submerged box culverts (2.4 x 2.4 x 44 m); two usually dry


box culverts (1.8 x 1.8 x 44 m); four cylindrical culverts (0.9 m in diameter x 44 m).

The rate of road-kill was investigated 12 months prior to the construction of the barrier wall-culvert system built along the stretch of highway in the Paynes Prairie State Preserve and was compared to the road-kill rate post-construction. The duration of each survey was 12 months (pre-construction Aug 1998-Aug 1999; post-construction March 2001-March 2002) with one sampling period scheduled for each week. Each sampling period consisted of 3 consecutive 24-hour sampling units. All road-kill was surveyed on the paved highway surface of both the north- and south-bound lanes, on the grassy shoulders (the right-of-way) and the median strip for the entire survey area. The length of the study area was 3.2 km, which included 400 m of road beyond the barrier wall-culvert system. During the same study period the use of culverts was also investigated with funnel traps, crayfish traps and/or infra-red cameras. Traps were sampled five nights per week. Cameras were monitored five days per week.

A total of 3101 dead vertebrates were recorded during the pre-construction phase of the survey, compared to 1891 dead vertebrates post-construction, with 68.8 per cent of all post-construction road-kill being hylid treefrogs. Hence, mortality was reduced by 65 per cent. Excluding hylid treefrogs, which were able to easily climb over the wall, mortality was reduced by 93.5 per cent after the construction of the barrier-wall system. Most post construction mortality was recorded in August and September and 73 per cent of road-kills occurred beyond the barrier-wall culvert system.

In addition to a decrease in mortality, there was an increase in the use of culverts. 51 vertebrate species were recorded to use the culverts during post-construction monitoring compared with 28 species in the 4 existing culverts prior to construction. The species recorded included fish, amphibians, reptiles and mammals.

Overall, the authors conclude that the barrier-wall system meet five of the six criteria outlined by Forman et al. (2003) to judge success as it is effective in reducing wildlife mortality especially of snakes, turtles and alligators, maintaining habitat connectivity, maintaining genetic interchange, allowing dispersal and recolonisation, and maintaining metapopulation processes and ecosystem function.


Singer, F. J. (1978) Behavior of mountain goats in relation to U.S. Highway 2, Glacier National Park, Montana, Journal of Wildlife Management, 42(3): 591-597

Location of study: Glacier National Park, Montana, USA

Habitat: Steep valleys 1030-1280 m in elevation, peaks 2133-3048 m elevation. Vegetation not described.

Target species: Mountain goats, Oreamnos americanus

Mitigation structure/s: One 2 m high x 3 m wide underpass, length not given

A behavioural study of mountain goats crossing the road to access a salt lick. This was compared to behaviour at a salt lick not affected by a road. The study was not focused specifically on mitigation structures, but the road included an underpass that was rarely used by mountain goats. Fencing was not used to funnel animals, or keep them off the road.

The routes taken by goats during 87 successful road-crossings (692 goats) and 31 unsuccessful crossings (101 goats) were observed in 1975. Of these, only nineteen goats attempted crossings via the underpass. It is suggested that the underpass was not effective because it was not located along the goat’s direct travel path, was too small, and approach from one side lacked vegetative cover.

Clevenger, A.P. & Waltho, N. (2005) Performance indices to identify attributes of highway crossing structures facilitating movement of large mammals, Biological Conservation, 121: 453-464.

Location of study: Banff National Park, Alberta, Canada

Habitat: Montane, sub-alpine and alpine vegetation, mountainous area, 1300 to over 3400 m elevation (from description in other papers).

Target species: Large mammals

Mitigation structure/s: Two creek bridge underpasses, 3 m high x 11 m wide. Five elliptical, metal culvert underpasses, 4 m high x 7 m wide. Four prefabricated concrete box underpasses, 2.5 m x 3 m. Two overpasses, 50 m wide.

Studied mitigation structures passing under or over an 18 km stretch of


the Trans-Canada Highway. The structures were specifically intended for the movement of wildlife. The use of structures by mammals was studied using track-pads, checked every 3-4 days over 34 months (1997-2000). Infra-red cameras were also installed at overpasses. Crossings were identified by tracks running in the same direction on trackpads at both ends of the structure.

The paper acknowledges a limited sample size (13 statistical replicates) and could not control for confounding variables in a large-scale ecosystem study. They address this by developing species specific performance indices and testing these against attributes of the crossing structures.

4209 large mammal crossings were observed, including bears (Ursus americanus, U. acrtos), wolves (Canis lupus), cougar (Puma concolor), elk (Cervus elaphus), deer (Odocoileus sp.) and humans. It was found that structural attributes best explained the performance indices for both predator and prey species, with landscape and human-related factors of secondary importance. This was a different result to earlier studies in the same area. The study suggests that more monitoring and multiple-species approaches are required to determine the effectiveness of mitigation structures. To maximise connectivity for large mammals a diversity of crossing structures, of mixed sizes are recommended. Continuous, long-term monitoring is also advocated to ensure structures remain functional.

Clevenger, A.P., Chruszcz, B. & Gunson, K. (2001) Drainage culverts as habitat linkages and factors affecting passage by mammals, Journal of Applied Ecology, 38: 1340-1349


Habitat: Montane, sub-alpine and alpine vegetation, mountainous area, 1300 to over 3400 m elevation

Target species: Small and medium-sized mammals

Mitigation structure/s: Thirty-six drainage culverts (dimensions not given)

Investigates the use of drainage culverts by mammals over a 24-km stretch of the Trans-Canada Highway. Surveys were conducted using sooted track-plates and checked at least twelve times over two winters (1999 and 2000). Tracks in the snow were also studied along transects


near the culverts to determine the presence of mammals in the area. A track was counted as a crossing if it occurred within a 20-m radius of culvert openings, even if there was no record on the track-plates.

The most common species recorded were weasels (Mustela erminea) and deer mice (Peromyscus maniculatus). Other species included bushy-tailed wood rats (Neotoma cinerea), American marten (Martesamericana), coyote (Canis latrans), snowshoe hare (Lepus americanus), red squirrel (Tamiasciurus hudsonicus), shrews (Sorex sp.) and voles (Arvicolinae spp). The number of crossings recorded was compared to that expected from the relative abundance in the area to get a performance index. Stepwise multiple regressions were conducted on the indices for small and medium-sized mammals.

Overall, the study suggests that road design and landscape features were more important than culvert design or dimensions for species use. Traffic volume was found to be highly important for all species. The dimensions of the tunnel influenced some species. The paper suggested that drainage culverts can mitigate harmful effects of roads, but acknowledges that crossing away from the structure and population-level impacts were not assessed.

More frequent culvert-intervals (150-300 m) and a mixture of culvert-sizes are recommended for the movement of wildlife. Vegetation cover near culvert entrances is also recommended.

Clevenger, A.P & Waltho, N. (2000) Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada, Conservation Biology, 14(1): 47-56


Habitat: Montane, sub-alpine and alpine vegetation, mountainous area, 1300 to over 3400 m elevation (from description in other papers).

Target species: Large mammals

Mitigation structure/s: Nine cement open-span underpasses and two metal culverts. Ranged in size from 4.2 - 14.9 m wide, 2.5 - 4.0 m high, and 25 - 91 m long.

Studied the crossing rates of large mammals through culverts under the Trans-Canada highway. Crossings were determined using track-pads at both ends of the culvert, with tracks in the same direction at both ends


indicating successful crossing. This was compared to crossing rates expected from relative abundance in the area, calculated using radio-tracking, scat counts and habitat suitability indices.

The species detected were grizzly bear (Ursus arctos), black bear (U. americanus), cougar (Puma concolor), wolf (Canis lupus), deer (Odocoileus sp.), elk (Cervus elaphus) and moose (Alces alces). For each species the number of crossings was compared to that expected from the relative abundance of animals in the area, giving species-performance ratios. These were compared to various attributes of the underpass using curvilinear and polynomial regressions.

The study suggests that the importance of structural attributes varies with species and scale. The influence of human activity is thought to be significant at all scales, and limiting human access and moving walking tracks are recommended. Overall, structural attributes were not good at explaining species performance indices. Carnivores preferred underpasses close to drainage systems, while ungulates avoided them (possibly due to a predator-prey interaction). Ungulates preferred long, narrow tunnels but this was interpreted to be a statistical artefact, due to correlations with noise and distance-to-town.

Valladares-Padua, C., Cullen, L. & Padua, S. (1995) A pole bridge to avoid primate road-kills, Neotropical Primates 3(1): 13-15.

Location of study: Morro do Diabo State Park, Sao Paulo, Brazil

Habitat: Fragmented forest. Details not given

Target species: Black lion tamarins Leontopithecus chrysopygus

Mitigation structure/s: One canopy, consisting of timber pole 8 m in length elevated 6 m above the road

Incidental observations of primates crossing the road. Endangered Black lion tamarins Leontopithecus chrysopygus were observed crossing a dirt road as part of a long-term study of the primate community in the area. This prompted the construction of a canopy bridge overpass, placed exactly in the location animals were previously observed crossing. The bridge, consisting of a horizontal timber pole suspended between two vertical timber poles, was constructed in mid-August 1991.

Although no detailed study of usage was conducted, black lion tamarins and capuchins (Cebus apella) were observed using the structure to


cross the road in both directions as soon as it was assembled. Between bridge construction and the end of 1994 at least 40 incidental observations have been made of these two species using the bridge.

It is thought that the bridge is being used constantly, probably daily, and that this has reduced the risk of road mortality. This is considered an important contribution to the protection of critically endangered Black lion tamarins. It is suggested that primates will avoid crossing on the ground where possible, making simple bridge designs appropriate.

Ng, S.J., Dole, J.W., Sauvajot, R.M., Riley, S.P.D. and Valone, T.J. (2004) Use of highway undercrossings by wildlife in southern California. Biological Conservation 115 (3): 499-507.

Location of study: near Los Angeles, California, USA

Habitat: a diversity of biological communities including chaparral vegetation (Adenostoma fasciculatum, Ceanothus spp., Rhamnus ilicifolia), coastal sage scrub (Artemisia californica, Salvia phyla, Malosma laurina), coast live oak (Quercus agrifolia) woodland and riparian woodlands (Salix lasiolepis, Platanus racemosa)

Target species: mammals, especially species of conservation concern including native carnivores and mule deer, and domestic animals

Mitigation structure/s: In total 15 underpasses: 4 termed spanning bridge underpass (surface roads or wide streams crossing under highway), 5 drainage culverts (square or pipe culverts) and 6 termed square culvert or tunnel (livestock tunnels). Dimensions of all structures given in detail.

Studied the use of 15 pre-existing underpasses by large carnivores, deer, mid-sized mammals and domestic mammals along three highways in California, USA over a period of 12 months (July 1999 – June 2000). Surveys were carried out over four consecutive days per month with remotely triggered cameras or gypsum powder track stations. Cameras and track stations were placed at each entrance and in the middle of the structure to detect crossings. Crossings were categorized as verified (evidence at both entrances and in the middle), probable (at both entrances but not in middle or at one entrance and in middle) or assessment of entrance only.

Predictor variables for animal use were also evaluated using Spearman's rank correlation. Variables included the dimensions of the structure,


human activity and the nature of the surrounding habitat (natural, developed or urban areas and landscaped).

2723 detections were recorded: 1640 (60.2 per cent) were of human activity (including horses, bicycles, and other vehicles), 521 (19.5 per cent) were native large and mid-sized mammals, 397 (14.6 per cent) were of small mammals and 155 (5.7 per cent) were domestic animals.

Raccoons Procyon lotor were the most commonly detected animal using the underpasses. This study also showed that large and mid-sized mammals (e.g. bobcat, coyote, domestic dog and cat) regularly use underpasses beneath highways even though these passages weren't purpose built for wildlife.

The use of underpasses by mule deer Odocoileus hemionus was significantly influenced by passage dimensions with all crossings occurring at large spanning bridge underpasses. Human activity significantly influenced the use of passages by domestic cats and dogs, and coyotes. The presence of natural habitat at the entrance of the passages influenced the crossing behaviour of deer, racoon and large carnivores in particular bobcats.

It was suggested that for the future installation of underpasses there should be consideration given to passage dimensions, especially if specifically built for deer. The likelihood of utilisation could be increased by installing animal-proof fencing to funnel wildlife to passages and by restoring habitat.

Olsson, M. P. O. & Widen, P. Effects of highway fencing and wildlife passages on space use and movements of moose Alcesalces in southwestern Sweden. Wildlife Biology (in press).

Location of study: E6 Highway, SW Sweden

Habitat: Mosaic of Spruce and Pine forest and farmland

Target species: Moose, Alces alces

Mitigation structure/s: Two overpasses (60 m long, 13 – 17 m wide with 2m high fencing to shield from oncoming traffic), one underpass ((35m long, 4.7m high x 13 m wide), all including gravel roads with low traffic (2 – 15 vehicles per hour).

This study documented the change in the spatial organisation of moose before, during and after the E6 (European Highway 6) was upgraded


from two-lanes to four-lanes with fencing. 24 Moose were fitted with GPS collars and locations recorded every two hours. Crossings were detected whenever fixes were taken on opposite sides of the road, and/or when recorded with camera and sand-pads within the structures. Moose are not a threatened species in Sweden and numbers have increased since the 1960s. Consequently, the recorded rate of collision with vehicles is high (at about 4000 annually), with some estimates of total accidents at around 8,000.

Rate of crossing decreased by 67 – 89 per cent after fencing, and all recorded crossings were over the two wildlife overpasses (none through the underpass). Number of moose mortalities due to collision with vehicles also reduced significantly. Fencing served to dramatically reduce mortalities, however it also served to significantly increase the barrier effect. The two overpasses were not sufficient to maintain the same rate of crossing as was recorded before fencing.

This paper quantified the movement of moose at one location before, during and after road duplication and fencing. The use of GPS technology AND tracking pads and cameras allowed for the identification of crossing rates at non-designated wildlife crossings.

Braden A., Lopez R. R. and Silvy N. J. (2005) 'Effectiveness of fencing, underpasses, and deer guards in reducing Key deer. mortality on the US 1 Corridor, Big Pine Key, Florida.' Texas A & M University.

Braden A. W., Lopez R. R., Roberts C. W., Silvy N. J., Owen C. B. and Frank P. A. (in press) Florida Key deer underpass use and movements along a highway corridor. Wildlife Biology.

Habitat: Big Pine Key, Florida – coastal mangroves and Buttonwood forest, grading into hardwood forest at slightly higher elevations.

Target species: Florida Key deer Odocoileus virginianus clavium

Road condition (after upgrade, if any): US Highway 1, 2-lanes, 18,000 vpd, 72 and 56 kmh speed limit during the day and night, respectively

Mitigation structure/s: 2 box underpasses (14 m x 8 m x 3 m) and 4 experimental deer guards (steel grids placed on the road that permit vehicles but prevent access by deer)


This study investigated the effect of road fencing and installation of mitigation structures on the movement, mortality due to collision with vehicles and rate of crossing of Florida Key Deer.

A total of 76 deer were radiotracked daily for two years before and one year after construction of the mitigation structures. Rates of mortality come from records collected since 1966 on all road-kill of Key Deer. Each underpass was fitted with a camera and in adjacent vegetation to estimate density in adjacent habitat.

Home range size did not change before or after mitigation. Before fencing, all relevant radiocollared animals (9 of 9) crossed the road, while after fencing only 5 of 11 crossed. Deer-vehicle collisions decreased by 94 per cent after fencing, with most entering the fenced road via deer guards, resulting in two mortalities. Underpasses were used a lot, with 2522 exposures (with one exposure presumably equalling one crossing?) during the first 12 months. Underpass use was significantly higher in the second 6-month period of monitoring (when compared to the first 6 months immediately after construction), suggesting an acclimation period.

The main conclusions were that fencing reduced the rate of road-kill, but increased barrier effects, and that the rate of crossing did not return to pre-fencing rates during the course of monitoring. Indeed, the authors imply that deer prefer to cross over the road rather than through the underpass because of differential rate of use of the two underpasses. The underpass with lower rates of use was nearer (ie < 200m) to the unfenced section of road, providing an alternate crossing location.

Pedevilliano, C. and Wright, R.G. (1987) The influence of visitors on mountain goat activities in Glacier National Park, Montana. Biological Conservation, 39: 1-11.

Location of study: Glacier National Park, Montana, USA

Habitat: Highway in deep canyon. Mineral lick, which is used by mountain goats, is on one side of the highway.

Target species: Mountain goat Oreamnos americanus

Mitigation structure/s: two underpasses: a specially designed goat bridge 4-9 m high, 13 m wide and 27 m long and a modified existing bridge spanning over creek and 55 m long. Retaining earth wall 4 m


high and 8 m wide and fence to restrict movement onto highway.

Investigated use of two highway underpasses by mountain goat to access a mineral lick during the summers of 1983 and 1984 (May - August) and compared the rate of crossing with pre-construction information. Also investigated the extent to which human activity (vehicular and visitor viewing) influences crossing behaviour. Observations of goat underpass use were made from a hide and the number, sex and age-class of goats and the time of day and direction of travel was recorded. Rate of traffic along highway, noise level in the underpass and activity by park visitors was simultaneously recorded. Park visitor activity was noted as either viewing animals from the highway shoulder over the underpass or from the purpose built observation platform 150 m from the lick. The corresponding behavioural response of the goats to visitor activity was recorded.

The underpasses and fences were considered effective in channelling goats under the highway. The pre-construction study found that goats successfully crossed the highway 74 per cent of the time compared with 100 per cent in the present study. There were fewer goats running back from the highway in the post-construction study (24 per cent) compared with the pre-construction (44 per cent) indicating that animals were happy to use underpasses rather than contend directly with vehicles. Also no highway mortality during the post-construction study was noted. Erect tails (as an indicator of disturbance or alarm in goats) were more prevalent in the pre- (70 per cent) rather than the post-construction study (55 per cent).

The rate of crossing differed through the day displaying crepuscular behaviour, which corresponds to the peak use of the mineral lick and when traffic rate was least. The pre-existing bridge had a significantly higher rate of crossings than the purpose-built bridge, possibly due to more room for crossing, greater visibility across the highway or due to the noise levels being higher in the purpose-built goat bridge than in the pre-existing bridge. There was a decrease in crossings as the number of vehicles per hour increased. As the number of visitors on highway shoulder increased, the longer it took for goats to successfully complete crossing. However vehicles caused more of a response than visitors on the highway.

The authors concluded that goats on the lick were not disturbed by people on the observation platform and is therefore sufficiently


designed and located. On the other hand, visitors that viewed the goats from the highway shoulder influenced the behaviour of goats crossing under the highway by causing run backs, hesitation and eliciting visual alarm responses.

They suggest that reiterating to visitors the fact that the highway shoulder is a no-parking zone and stopping is dangerous to human safety may enhance crossing behaviour. The installation of an additional viewing platform may lessen the incentive to stop and that stress levels in goats may be reduced by screening the underpasses with vegetation and by providing an acoustic buffer.

Yanes, M., Velasco, J.M. and Suárez (1995) Permeability of roads and railways to vertebrates: importance ofculverts. Biological Conservation. 71: 217-222.

Location of study: Central Iberian Peninsula, Madrid region, Spain

Habitat: pasture, pastures with maquis and/or scattered woodland, and forest or maquis with scattered woodland.

Target species: vertebrates. Species detected: Amphibians, reptiles, small mammals, lagomorphs, small carnivores, canids

Mitigation structure/s: 17 pre-existing culverts: 2 under motorway; 5 under railway line; and 10 under several local roads. Mean measurements given: 1.2 m high x 1.2 m wide x 13.1 m long, measurements of each culvert not given. No fencing to funnel fauna into culverts. Some had boundary fences parallel to road/rail.

Investigated the use of 17 pre-existing culverts under a motorway, several local roads and a railway line. Over the course of twelve months, a total of sixteen days (four days per season, eight in areas with wild ungulates) were sampled by means of marble dust to identify vertebrate tracks. A dust station was placed in the middle of each culvert and in summer and winter 4 additional dust stations were placed along the roadside: 2 2-4 m from culvert (one each side of road), 2 40-50 m from culvert (one each side of road). To determine the characteristics of the culverts that encouraged their use ten variables were recorded. These included 4 culvert dimensions (length, height, width and openness), 3 road/rail characteristics (width used by traffic, total width and height of boundary fence), traffic intensity, vegetation complexity, and the presence/absence of detritus pits. Data analysis


included a calculation of an index based on the number tracks found inside and outside the culverts; Spearman’s rank correlation to investigate the relationship between use and culvert dimensions and characteristics; and Kruskal-Walis and Mann-Whitney U-tests to investigate the relationship between use and traffic, vegetation and detritus.

The most common species to use the culverts were mice Apodemus sylvaticus, shrews Sorex spp. and rabbits Oryctolagus cuniculus. Reptile and small mammal tracks were more abundant inside the culverts than on the outside dust stations. Culverts have soil and debris covering the ground surface, presumably providing a less hostile environment than the road/rail surface. There was no statistical difference in the number of rabbit and carnivore tracks found in the culverts and outside. Most crossings occurred during spring and summer. Carnivores and small mammals showed least seasonality. Reptiles used culverts significantly more in summer than the rest of the year which coincides with other studies that found an increase in activity during warmer months. Seasonal variations in culvert use were generally related to the abundance of animals in surrounding habitat.

The characteristics of the culverts that influenced use varied between groups. The dimensions of culverts appeared to affect small mammals, as well as road width and vegetation complexity. The crossing index of medium-sized mammals (rabbits and carnivores) was affected by the total width of the road. Rabbits were also negatively influenced by the height of the boundary fence. The fences do not allow access to culverts and therefore hinder their use. Consideration is needed to construct fences that both reduce animal access to the road but direct fauna towards crossings. The presence of detritus pits was related negatively with the passage by reptiles and rabbits or carnivores did not use those culverts with detritus pits.

van Leeuwen, B.H. (1982) Protecting of migrating common toad (Bufo bufo) against car traffic in The Netherlands. Environmental Conservation. 9 (1): 34.

Location of study: The Netherlands

Habitat: not described

Target species: common toad Bufo bufo


Mitigation structure/s: eleven actions were used including the construction of a tunnel in combination with fences, pitfall traps to collect toads and carried over the road. Detailed description of mitigation not given.

This paper is an extremely brief summary of 11 actions used to prevent common toad being killed by traffic, especially during migration periods. Pitfalls were used in conjunction with a fence to collect toads, after which they were carried across the road. Another action involved a tunnel with fences. In other cases the road was temporary closed to traffic (three cases) or toads were collected by hand (four cases). The details of each action are not given.

The paper concludes that the actions are a success in reducing mortality (although there is mention of which action was more effective). Mortality of toads, before the actions were enforced, was 15 to 50 per cent of all migrating toads. After the actions had been taken, mortality was reduced to 5 per cent.

Mathiasen, R. and Madsen, A.B. (2000). Infrared video-monitoring of mammals at a fauna underpass. International Journal of Mammalian Biology. 65: 59-61.

Location of study: Highway E45 between Aarhus and Randers in Jutland, Denmark.

Habitat: Not given

Target species: mammals. Species detected: red fox Vulpes vulpes, badger Meles meles, stone martens Martes foina and roe deer Capreolus capreolus.

Mitigation structure/s: one underpass with entrance shaped like a half circle and is 13 m wide, 7.5 m high and 155 m long. There is a fence along the highway, which is 1.75 m high and extends 1 km either side of underpass.

Investigated the use of one underpass by mammals and observed the behaviour of animals as they entered the underpass. Surveys were carried out in two 30 day periods in April/May and August/September 1997 using an infrared camera. The underpass was under surveillance for 495 hours. Sand stations were also positioned at each end of the underpass.


In total there were 122 red foxes, 16 badgers, 18 stone martens and 20 roe deer recordings. Based on fur and behaviour at least three individual foxes, three individual badgers, one or two stone marten and one male roe deer were identified. The number of recorded crossings was higher in the April/May sampling period than in August/September.

The behaviour of the carnivores appeared not to change at the entrance of the underpass. On the other hand the roe deer was hesitant at the entrance often changed its behaviour by looking around, lowering its head and changing its manner of walking. Whilst in the underpass, the roe deer were observed to keep about 3.5 m from the wall. Galloping through the underpass was only observed in April/May. Brown hares Lepus europaeus were observed close to the entrance but never entered. Overall the authors suggest that the roe deer were hesitant possibly due to the size of the opening. The lack of galloping in August/September was suggested as being a result of the animals becoming habituated to the underpass during summer with heightened territorial defence.

Cain, A.T., Tuovila, V.R., Hewitt, D.G. and Tewes, M.E. (2003). Effects of a highway and mitigation projects on bobcats in Southern Texas. Biological Conservation. 114: 189-197.

Location of study: between George West and the southern boundary of Live Oak county, Texas, USA.

Habitat: typical of Rio Grande Plains including mesquite Prosopisglandulosa, blackbrush Acacia rigidula, cenizo Leucophyllum frutescens, prickly pear Opuntia sp., and live oak Quercus virginiana. There are also patchy thornshrub communities.

Target species: bobcat Lynx rufus

Mitigation structure/s: 18 crossing structures were monitored which including 4 bridges, 14 culverts, 5 of which were modified to encourage felid use. Modifications including raised catwalks to allow use during periods of heavy rainfall, and open medians and drop inlets to provide more light, air movement and decrease the tunnel effect. Net-wire fence 1.6 m high, extending 100 m either side of culvert, on both sides of highway. Fence installed one year after study commencement, paired assessment of fence effectiveness: 6 with fence, 6 without.

Animals were trapped between June 1997 and April 1999 along the


edge of the highway and 16 were fitted with a radio collar. Each radiocollared animal was located 12-15 times per month from July 1997 to May 1999. Radiotracking data was used to assess if bobcats were preferentially using certain cover types. Surveys of bobcat mortalities were conducted between June 1997 and May 1999 by driving the north and south bound lanes 1-2 time per week. This information was used to determine if mortalities occurred disproportionately in certain cover types.

The 18 culverts were monitored by means of cameras and track stations. Cameras were installed at both entrances of seven culverts. Cameras were rotated every 2 weeks to ensure all crossing structures were monitored equally during the study period (Aug 1997 to May 1999). All 18 structures were monitored using track surveys. Stations were checked twice each week. All cats were combined into a felid category because it was difficult to differentiate between species. The following 5 variables were measured to explain the use of the culverts: openness (width x height/length); water (per cent of days each month culvert contained standing water); visibility (if animal standing at entrance could see completely through crossing or not); cover (per cent canopy cover of woody vegetation �100 m of structure); and distance to nearest thornshrub habitat. 12 culverts were paired based on similar dimensions and crossing use indices (results from first year of study) to test the effectiveness of fencing. A fence was installed at one culvert from each pair.

There were felid tracks detected in all three types of crossing structures, however crossing use was higher in bridges and modified culverts than unmodified culverts. Modification of culverts was considered successful in increasing the likelihood that bobcats would use the culverts. There was no difference in the use of crossing structures across seasons. Culvert use was positively correlated to the openness ratio of the culvert and cover of woody vegetation near entrance. NB. no mention of each structure size so not sure how openness differs between structures e.g. bridges may have greater openness ratio than unmodified culverts, etc. This study was unable to provide optimal culvert dimensions but found that the minimum size of high use culverts was 0.9 x 1.6 m.

86 per cent of the photographs taken with the cameras were of bobcats, and 14 per cent were of feral house cats. The cameras recorded 54 complete crossings through culverts, 41 of which occurred during


darkness. Culverts were also used for other purposes such as resting, hunting, and perhaps thermoregulation. Fence construction had no significant effect on use by felids, however if only considering high use culverts, there was an indication that the fence increased bobcat use.

Radiocollared bobcats showed selection for thornshrub vegetation type over more open cover types. Vehicle caused mortality was more likely to be adjacent to habitat preferred by bobcats and in sections of the highway in which thornshrub occurred in the median. Thus maintaining thornshrub vegetation near roads to help reduce the effects of habitat fragmentation may increase the chance of mortality. Other techniques to reduce mortality, such as long fauna exclusion fences, and reduce fragmentation, such as building sufficient crossing structures adjacent to vegetation cover types preferred by bobcats, must be considered.

Rodriguez, A., Crema, G. and Delibes, M. (1996). Use of non-wildlife passages across a high speed railway by terrestrial vertebrates. Journal of Applied Ecology. 33: 1527-1540.

Location of study: mountainous area: Montes de Toledo, high speed railway between Madrid and Seville, Spain

Habitat: the valleys and foothills are cultivated with cereal, slopes and top of hills (up to 1300 m) are covered with scrub (mainly Cistus ladanifer), scattered trees (mainly Quercus rotundifolia), pine stands Pinus pinaster) and pasture.

Target species: vertebrates. Vertebrate groups detected: carnivores, lagomorphs, small mammals and reptiles.

Mitigation structure/s: 17 pre-existing structures (not purpose built for wildlife), 2 of which are flyovers, the rest were culverts and underpasses. Fence (2 m high topped with two strands of barb wire) was built part way through study.

Investigated the use of 17 non-wildlife passages by terrestrial vertebrate between September 1991 and July 1992. The study had four aims: 1) to observe which groups of terrestrial vertebrates use structures; 2) if vertebrates use occasionally or frequently (so rate of crossing); 3) to test the effect of several passage features on crossing rates; and 4) to test the effectiveness of fencing. Passages were monitored 15 – 22 days per month by using sand near one entrance. Presence/absence data used as estimates of crossing rate. The monthly


crossing rate for each passage = ratio of days with animal tracks/number of operative days. For each passage, three variables were recorded: 1) physical characteristics (length, width, height, presence/absence of pit); 2) distribution of cover (presence/absence of cover within 20 m of entrance and distance to scrubland patch); and 3) degree of human disturbance (distance to nearest house and monthly rate of use by humans).

Ungulates that were known to be present in the study area were not detected in the passages. The authors suggest that this may be due to narrow passage size and the fact that vehicles, livestock, people or dogs used the passages greater than 3 m in width almost daily. Hence, ungulates require specifically designed passages that take into account distribution of cover, passage dimensions, placement and human disturbance. Small mammals (55.6 per cent of records), carnivores (25.2 per cent), reptiles (10.7 per cent) and lagomorphs (8.5 per cent) were recorded to use the passages frequently. Fencing did not significantly influence relative crossing rates but the study was confounded by time

Crossing rates varied with habitat type and season. The presence of cover near one or both entrances favoured the use of passages by carnivores. Seasonal variation in use (highest use in summer and lowest in late winter) by carnivores coincides with expected changes in abundance (litters start to leave the den in early summer) and mobility (both dispersal and mating take place in autumn and winter). Low carnivore crossing rates were associated with the presence of pits.

Small mammals use of the passages peaked in late spring and summer, which corresponds with a tendency of spring births for most species in the temperate zone and an increase in deaths through autumn and winter. Passage use by this group was also significantly influenced by habitat with more crossings in the border than in the scrubland and farmland, and by culverts with small cross-sections.

Reptile use of passages peaked in late spring and summer, which corresponds with expected seasonal fluctuations in activity. Reptiles preferred culverts and underpasses rather than flyovers with culverts of intermediate width having the highest rate of crossing.

Rodriguez, A., Crema, G. and Delibes, M. (1997). Factors affecting crossing of red foxes and wildcats through non-wildlife


passages across a high-speed railway. Ecography 20: 287-294.

Location of study: eastern Montes de Toledo, high speed railway between Madrid and Seville, Spain

Habitat: hilly area with gentle slopes and continental climate. Cereal crops occupy valleys, hills covered with Mediterranean scrubland.

Target species: red fox Vulpes vulpes and wildcat Felis silvestris

Mitigation structure/s: 17 non-wildlife crossing structures: 12 culverts (including 7 culv 1.2 m diameter, 2 culv 2 m diameter and 3 culv 2 m wide which have deposition pits at one entrance), 3 underpasses (3.5 m wide, 3.5 m high) and 2 overpasses (6 m wide with wooded walls 2 m high). Passage length varied between 13 and 46 m.

A study investigating red fox and wildcat use of 17 non-wildlife passages under a high speed railway between September 1991 and July 1992. The study also investigated the influence of passage features on the frequency of crossing, especially the amount of habitat cover at the passage entrance and the level of human activity, and to investigate the effectiveness of a fence. Each passage was monitored for 15-22 days per month by using sand (records animal tracks) near one entrance. Presence/absence data used as estimates of crossing rate. The monthly crossing rate for each passage = ratio of days with animal tracks/number of operative days.

Each passage was assigned to one of three types of habitat (scrubland, border and open farmland). Then for each passage, five variables were recorded: 1) physical characteristics i.e. passage type (length, width, height, presence/absence of pit); 2) presence/absence of cover within 20 m of at least one entrance; 3) distance to nearest scrubland patch; and 4) distance to nearest inhabited farm and; 5) monthly rate of use by humans (calculated same way as monthly rate of crossing by fox and wildcat).

Highest total rates were recorded in the same passage for the fox (20.8 per cent) and wildcat (22.2 per cent). Habitat and month had significant effects on fox and wildcat crossing rates. The average monthly crossing rate of both fox and wildcats in the scrubland were 7 to 30 times higher than in the border and farmland. Observations from other studies found that the fox had a preference of scrubland for bedding. In addition the border habitat had high human activity

There were also temporal differences in fox crossing rates with more


crossings occurring between July and November. There were also temporal differences in wildcat crossing rates but with less marked differences than fox. The fluctuations in fox crossings from July to November coincide with expected breeding season and the dispersal season in the Iberian Peninsula. In both species there is a second peak in the rate of crossing (January for fox and December/January for wildcat) which appear to relate to the expected increase in mobility during the mating period.

The only passage feature that had a significant effect on fox and wildcat crossing rates was the presence of cover at the passage entrances. From the records of the direction of tracks, the fox and wildcat tended to enter passages by the entrance with cover.

Fox and wildcat crossing rates did not significantly differ before and after fencing. Therefore fencing might not be efficient for channelling fox or wildcat movement into passages

From this study the authors recommend that passages be suitably placement near (and preferably within) forested tracts and in open areas increase the cover near passages entrances.

Eide, S.H., Miller S.D. and Chihuly, M.A. (1986). Oil pipeline crossing sites utilised in winter by moose, Alces alces, and caribou, Rangifer tarandus, in southcentral Alaska. Canadianfield-naturalist 100 (2): 197-207.

Location of study: 145 km segment of Trans-Alaska Pipeline corridor from Meiers Lake to Squirrel Creek, southcentral Alaska.

Habitat: mixture of black spruce Picea mariana, white spruce Picea glauca, aspen Populus tremuloides, balsam poplar Populus balsamifera, paper birch Betula papyrifera, shrub birch B. glanulosa and willow Salix spp., interspersed with sedge meadows, shallow lakes and riparian habitats. Understories of Vaccinium spp., Ledum spp., lichens and mosses.

Target species: moose Alces alces and caribou, Rangifer tarandus

Mitigation structures: three types of crossings: specially-designed crossings where the pipe was elevated termed ‘DBGC’ (exceeded 3 m in vertical clearance, total number of crossings 81), long buried segments of the pipeline (average length 1.1 km) and sagbends (short buried segments length less than 18.3 m).


Explored the use of features of the pipeline that permitted crossing by moose and caribou, and determined which characteristics influenced use. Surveys were conducted over winter 1977/1978 between October and April. 145 km of pipeline was divided into three segments. Each segment was searched for tracks in the freshly fallen snow for 16-23 days during the study period. Data were collected on: 1. vertical clearance (between ground and bottom of pipe), location, 3. complete crossing or deflection (stop and turned around), 4. crossing at buried sections, sagbend or elevated section, 5. direction of travel.

Found that moose selection of crossing sites was not influenced by pipeline characteristics (elevated or buried sections) and moose did not select for the specially designed crossing sites (DBGC). However, lowest measured vertical clearance was 1.3 m. 3.2 per cent of moose encounters with the pipeline were deflections, occurring where vertical clearance was <3.4 m.

For caribou, found that most (70.3 per cent) crossings were eastward, relating to the fall migration. Caribou showed negative selection for low vertical clearances (less than 2.1 m). Showed selection for buried sections was highly significant. However, selection for pipeline characteristics hard to establish, as the crossing types were not randomly distributed. Eg. Buried segments were intentionally located in areas of historic caribou movements. Caribou crossing buried pipe may indicate continued use of traditional migration routes rather than active selection of buried segments.

Since the construction of the pipeline the caribou annual migration pattern had not changed. The crossing sites are therefore considered adequate in assuring free passage of caribou (and moose) across the pipeline.

Ascens�o, F. and Mira, A. (2007). Factors affecting the culvert use by vertebrates along two stretches of road in southern Portugal. Ecological Research. 22: 57-66.

Location of study: roads M370 and IP2, southern Portugal

Habitat: agro-forest - cork and holm oak tree stands (Quercus suber and Q. rotundifolia), open land such as pastures, meadows or extensive agriculture, and olive groves.

Target species: vertebrates


Mitigation structure/s: 34 culverts along two roads: 17 culverts per road. Exclusionary fence along IP2, 1.5 m high.

Evaluated the use of 34 culverts under 2 roads in southern Portugal and what characteristics influenced use. Track stations were positioned at both entrances of culverts to record vertebrate use. Sampling was carried out during three seasons (spring, summer and Autumn 2004). Each sampling period consisted of four operative days (tracks were checked every second day). Recorded presence/absence of species and from this calculated crossing rates for each culvert (ratio between number of days with tracks and number of operative days). Explanatory variables in the analyses including culvert design attributes (culvert width, openness, presence of detritus pit, vegetation cover at entrances), landscape attributes (land cover 300 m radius from culvert, average distance to adjacent culvert, underpass or overpass, distance to urban areas), and road attributes (traffic volume, presence/absence of fence). NB. IP2 had higher traffic volume and had fence.

Thirteen taxa were detected which including amphibians, reptiles, lagomorphs, small mammals, carnivores and domestic cats and dogs. The most commonly recorded species were stone martens, followed by Egyptian mongooses and genets. Vertebrate crossings were mainly observed during spring. There were significant differences in culvert use of the two roads by small mammals, hedgehogs, badgers, Egyptian mongooses, foxes and dogs. Small mammal and hedgehog crossing rates were higher along M370 whereas the crossing rates for the remaining animals were higher at IP2.

Where there was no cover at the culvert entrances, crossing rates were significantly higher for lagomorphs, badgers, foxes and dogs. Hedgehogs, lagomorphs, badgers, weasels, otters, foxes, cats and dogs were not detected where detritus pits were present. Generally use of culverts with respect to different design attributes was correlated with animal body size. For example, small mammals and reptiles selected for culverts with smaller length, lagomorphs selected for narrow passages. In terms of culvert location: culverts near open land were avoided by typical forest species such as genet and stone martens whereas lagomorphs, foxes, badgers and dogs preferred those culverts; and those culverts closest to urban areas were favoured by domestic cats and hedgehogs.


Guyot, G. and Clobert, J. (1997). Conservation measures for a population of Hermann's tortoise Testudo hermanni in southern France bisected by a major highway. Biological Conservation. 79: 251-256.

Location of study: Motorway A57, Cannet de Maures, southern France.

Habitat: Garrigue composed mainly of Mediterranean dwarf shrubs Cistus monspelvensis and Erica scoparia with a few cork oaks Quercus suber and evergreen oaks Quercus ilex.

Target species: Hermann’s Tortoise Testudo hermanii

Mitigation structure/s: 1 tunnel, 2 culverts. Fence on both sides of the highway, extending for 4 km.

The main focus of this study was to monitor the survival and movement of tortoise following the construction of a highway. Culverts, a tunnel and a fence were installed to alleviate the possible negative effects of the new highway. Animals were kept in captivity during the construction then each tortoise was individually marked and released in October 1990. In 1993 and 1994 marked individuals were searched for to estimate survival and the distance moved from the release point.

Seven individuals were recorded on the opposite side of the highway from the release point. The authors assume that these individuals used the culverts/tunnel to cross the highway. However there is no evidence to confirm this as the crossing structures were not monitored during the study period.

Only five individuals were killed on the part of the highway that was protected by fences, thus, the fence is believed to aid in keeping the road mortality low.

Aresco, M.J. (2005). Mitigation measures to reduce highway mortality of turtles and other herpetofauna at a north Florida lake. Journal of Wildlife Management, 69: 549-560.

Location of study: US Highway 27, Lake Jackson, northwestern Florida, USA

Habitat: 1,620 ha lake that has been divided by highway,

Target species: herpetofauna


Mitigation structure/s: one drainage culvert (3.5 m in diameter, 46.6 m long), connecting two portions of lake. Constructed fence along 700 m of Highway, designed to direct fauna into culvert. Bottom edge of fence was buried ca. 20 cm and above ground height was ca. 40 cm. The fence ran parallel to the highway and the ends were turned towards the lake for 80 – 100 m.

Calculated rate of mortality over a 40-day period. Built fence to direct fauna into existing culvert and evaluated the effectiveness of fence in reducing road mortality and facilitate migration over 2.5 years. Area adjacent to fence was monitored for migrating animals 2-4 times daily from 2000 to 2003. During the dry periods tracks in culvert were recorded. Turtles were marked to estimate total number of individuals crossing the highway. Compared number of dead and alive animals found on road before and after fence construction. Estimated the probability of turtle being killed in one crossing attempt.

In 44 consecutive months, recorded 10,229 reptiles and amphibians of 44 species (8,842 turtles of 10 species). Prior to construction, found 343 dead turtles on road in 40 days. During comparable 40-day period post fence construction, 6 dead turtles were found. Based on 2001 traffic volume, probability model predicted that only 2 per cent of turtles would successfully cross the highway. Therefore, the author states that at least 98 per cent of turtles diverted by fences probably would have been killed if fences were not in place. The following species were found moving through the culvert: turtles (Trachemys scripta, Pseudemys floridana and Apalone ferox), American alligator Alligator mississippiensis, and frogs (Rana sphenocephala, R. grylio, R. catesbeiana) and cornsnake Elaphe guttata. Recorded tracks of > 200 individual turtles.

The author concluded that the fence in combination with an existing culvert was effective in reducing road mortality and facilitate highway crossings. The installation of an extra fence at the ends (this turned away from the highway for 80-100 m) meant that animals were not able to access the highway from around the fence. Road mortality rate was significantly less after fences were installed and turtles were observed using the culvert. Few animals (only <1 per cent) were able to access the highway by climbing or penetrating the fences. The intensive patrolling of the fence during the study period supposedly reduced the trespass rate. Thus fences would have to be monitored daily or the fence redesigned to disallow turtles to climb the fence.


Lehnert, M.E. and Bissonette, J.A. (1997). Effectiveness of highway crosswalk structures at reducing deer-vehicle collisions. Wildlife Society Bulletin. 25: 809-818.

Location of study: three roads (SR 32, SR 248 and US 40) in Summit and Wasatch counties, northeastern Utah, USA

Habitat: Oakbush Quercus gambelii and sagebrush Artemisia spp. grass communities along drainage slopes and foothills. Riparian communities dominated by cottonwood Populus angustifolia trees and willow Salix spp. thickets, on lower valley areas along the margins of the Provo River. Pastureland was also common throughout lower valley regions.

Target species: mule deer Odocoileus hemionus

Mitigation structure/s: 8 crosswalks (guided walkway on road) with fences (2.3 m high) to funnel animals onto crosswalk. Also installed 3 warning signs for traffic

Aim of study was to evaluate the effectiveness of crosswalk to allow highway crossings whilst reducing deer-vehicle collisions. Highway mortality was monitored at least once a week at experimental and control sites before and after construction of crosswalks (prior to construction Oct 1991-Sept 1994, post construction Sept 1994 - Nov 1995). Monitored highway mortality in experimental and control areas before and after the construction of crosswalks. Use of the crosswalks was determined by spotlighting twice monthly. Deer behaviour and movement patterns in crosswalk zones were assessed with night-vision goggles. Also conducted motor-vehicle speed assessments to evaluate motorist response to crosswalk warning signs.

At highway US 40 (four lane highway) had an experimental and control sections. Mortality at these sites during 36 months prior to construction found 148 and 123 deer at the experimental and control section, respectively. Post construction found 63 deer at control and 46 deer at experimental, hence suggesting that there was a 42.3 per cent reduction in highway mortality following the construction of the crosswalks. Prior to construction found 111 killed deer at the experimental road SR 248, and 75 deer at control road SR 32. Post construction found 36.8 per cent reduction in road mortality (32 at control, 36 at experimental). However no statistical evidence that


observed mortality reductions were a result of mitigative efforts.

Recorded 9 instances in which deer entered crosswalks to attempt road crossing. First three deer to cross went outside the confines of the crosswalk and became trapped on the road between the fences. Vehicles along US 40 killed two of these animals. Deer in the crosswalk were more likely to walk on the road surface than deer at the control sites. Fencing and crosswalks did not appear to reduce the tendency for deer to use the right-of-way for foraging. Deer travelled faster on highway US 40 crosswalks than on SR 248 crosswalks. The crosswalks did not appear to disrupt seasonal deer movements. The 1-way right-of-way escape gates were ineffective in allowing trapped deer on the right-of-way to move away from the road.

Motorists did not reduce speed while passing through the crosswalk zones. The success of the mitigation system depends heavily upon motorists reducing speed in designated crossing zones, thus other forms of warning signs, such as flashing lights or the use of pavement, is suggested.

Gloyne, C.C. and Clevenger, A.P. (2001). Cougar Puma concolor use of wildlife crossing structures on Trans-Canada highway in Banff National Park, Alberta. Wildlife Biology 7: 117-124.

Location of study: Trans-Canada Highway, Banff National Park, Alberta, Canada

Habitat: montane and subalpine habitats: forests dominated by Douglas fir Pseudotsuga menziesii, white spruce Picea glauca, lodgepole pine Pinus contorta, aspen Populus tremuloides and natural grasslands

Target species: cougar Puma concolor

Mitigation structure/s: Five structure types among 22 crossing structures: seven open-span bridge underpasses (3 m high, 11 m wide), three creek bridge underpasses (open-span bridge underpasses with running water), six metal culvert underpasses (4 m high, 7 m wide), four concrete box culvert type underpasses (2.5 m high, 3 m wide), and two 50 m wide overpasses. Mitigation measures were constructed in phase 3A which was completed in late 1997.

Investigated the use of wildlife crossing structures by cougar between November 1996 and July 2000. Main questions of study included: whether there was a seasonal pattern in use, a correlation between use


by cougars and main prey (white-tailed deer Odocoileus virginianus and mule deer O. hemionus), a correlation between use by cougars and humans, and whether cougars selected for a particular type of structure. Track stations were set at both ends of crossing structure and checked at 3-4 day intervals. Cameras were installed at the overpasses as a supplement to the track stations.

Cougars used the crossing structures more during winter than summer. Cougar and human use were not correlated. This may be due to the fact that cougars are crepuscular (mainly active at dawn and dusk) and human use of the structures is almost exclusively diurnal. There was a positive correlation between the use of the structures by cougars and that of white-tailed and mule deer. Deer are a key component of the cougar diet and hence cougars will select a home range based on the presence of deer. The crossing structures that were used most by cougars were those situated close to high quality cougar habitat.

Open-span bridge underpasses (58 per cent of cougar passages) were used more than the other wildlife crossing structures. Creek bridge underpasses were used in proportion to their availability. For all other structure types use was significantly less than expected. In conclusion the authors suggest that whilst planning mitigation it is best to invest funds in underpasses rather in the costly overpasses. However overpasses may be used in time as vegetation becomes dense and provides overhead cover. Three cougars were killed on a fenced, mitigated section relatively close to crossing structures. The authors highlight the need for a more effective fence design.

Reed, D. F. (1981) Mule deer behavior at a highway underpass exit, Journal of Wildlife Management, 45(2): 542-543; and

Reed, D. F., Woodard, T. N. & Pojar, T. M. (1975) Behavioral response of mule deer to a highway underpass

Location of studies: Colorado, USA

Habitat: Pine/aspen forest and more open shrubland. Topography is steep, with valleys at 1030 to 1280 m elevation, and peaks/ridges at 2133 to 3048 m elevation.

Target species: Mule deer, Odocoileus hemionus

Mitigation structure/s: One concrete box underpass 3.05 x 3.05 m wide and 30.48 m long.


Behavioural studies of deer moving through a single underpass under a 4-lane, interstate highway. The underpass was constructed along known deer migration trails, but also aids drainage. Fencing was used to keep animals off the road.

The original study monitored the underpass during migration periods over 3 years using overnight time-lapse video recording (1970-1973). Movements were also recorded using electro-optical counters and morning track counts and direction. The behaviour of deer exiting the underpass was observed from the video-recording. The second study involved direct observation of deer at the underpass over another six years (1974-1979).

The studies suggest that deer are reluctant to move through the underpass and recommend larger underpasses with at least 4.27-m height and width, and minimal length. They also recommend dirt floors, small skylights and suggest that artificial lighting was unnecessary. The conclusions are drawn from looking at a single underpass, with no controls used.

Van Wierwen, S.E. and Worm, P.B. (2001). The use of a motorway wildlife overpass by large mammals. Netherlands Journal of Zoology. 51 (1): 97-105.

Location of study: A50 motorway at Terlet, between Arnhem and Apeldoorn, Netherlands

Habitat: a large area (>10,000 ha) of forests and heathlands

Target species: mammals

Mitigation structure/s: 'Terlet' overpass (50 m wide x 95 m long), built at ground level. Fences were erected to prevent animal access to the road

The use of the 'Terlet' overpass by red deer Cervus elaphus, wild boar Sus scrofa and roe deer Capreolus capreolus was investigated with sand spanning the width of the overpass. In 1989, shortly after being built, there was a year round study to estimate the use of 'Terlet'. The study was repeated between May 1994 and April 1995. In 1989, sampling frequency was twice weekly. During the 1994/1995 study sampling was less rigid to avoid adverse weather. A short study was carried out in 1995 to record the use by mice and other small mammals. Traps were set on the overpass during 5 summer nights.


In 1989 red deer, wild boar, roe deer and red fox Vulpes vulpes were recorded using the overpass. The 1994/1995 data collection recorded fallow deer Dama dama, Badger Meles meles and highland cattle Bos taurus were recorded in addition to those found using the overpass in 1989, indicating familiarity of the overpass by these animals. In 1994/1995 there was an almost threefold increase in red deer crossings compared with crossings recorded in 1989 even though overall density remained the same (unsure how density established). This indicates that the red deer become more habituated to the overpass. The badger was using the overpass as a feeding ground. The authors speculated that Highland cattle dung attracts many beetle which are a stable part of the badger diet.

Wild boar and red deer used the overpass less frequently from late winter to early summer. This coincides with the expected rutting season with the majority of red deer prints at that time being from solitary males.

Trapping revealed that wood mice Apodemus sylvaticus, common voles Microtus arvalis and common shrews Sorex araneus use the overpass. There was a difference in the catch rate with the highest on the northern slope where there is vegetation that provides adequate shelter compared with the southern slope that has oak trees and bare soil.

The authors conclude that overpasses can be an effective means of connecting habitat.

Mata, C., Hervás, I., Herranz, J., Suárez, F. and Malo, J.E. (2005). Complemantary use by vertebrates of crossing structures along a fenced Spanish motorway. Biological Conservation. 124: 397-405.

Location of study:

Habitat: 20 km of non-irrigated arable crops interspersed with patches of scrubby woodland dominated by holm oak Quercus rotundifolia and gum cistus Cistus ladanifer. 30 km mosaic of sub-oceanic holm oak woods, patches of gum cistus and leguminous shrubs Cytisusmultiflorus, C. scoparius and Agrostis castellana pastures with a small proportion of cultivated fields. The remaining section of Pyrenean oak Quercus pyrenaica woodland, tall Cytisus spp. and Adenocarpus complicates scrub, low scrub (Genista tridentate, Halimium ocymoides,


H. lasianthum) and humid pastures.

Target species: terrestrial vertebrates

Mitigation structure/s: 82 crossing structure: 33 circular culverts (for drainage: 1.8 m diameter, length 36-80 m), 10 adapted culverts (drainage but adapted for wildlife: 2-3 m wide, 2 m high, 35-50 m long), 14 open span underpasses (for rural tracks and livestock: 4-9 m wide, 4-6 m high, 34-66 m long), 7 wildlife underpasses (purpose-built for wildlife: 20 m wide, 5-7 m high, 30-36 m long), 16 overpasses (rural tracks: 7-8 m wide, 58-62 m long), and 2 wildlife overpasses (purpose-built for wildlife: 16 m wide, 60 m long). Fence along entire length of motorway.

Objectives of study were to investigate the use of 82 crossing structures by terrestrial vertebrates between the end of June and beginning of September 2002 and to analyse which features of the structure, surrounding habitat and human activity influence the use of these structures. Each passage was monitored with a marble dust track station placed at the half waypoint of the structure until there were 10 days worth of valid records. To compliment the tracking stations, infrared cameras were installed in 47 of the structures. Study used multivariate analysis.

A total of 17 species and taxonomic groups were recorded using the marble dust with small mammals being the most frequently detected, followed by lagomorphs, large canids and red fox Vulpes vulpes. The simultaneous study using cameras separated lagomorphs into rabbit Oryctolagus cuniculus and Iberian hare Lepus granatensis, and large canids into dogs Canis familiaris and 1 record of a wolf C. lupus. Almost all of the mammal species present in the area were detected using the structures (only roe deer, wild boar and otter were not recorded).

As found in other studies, generally there was a direct relationship between animal size and the dimensions of the crossing structure. Small mustelids, amphibians, reptiles and small mammals were using the circular and adapted culverts more often than the other structures, whereas lagomorphs, red fox and large canids were preferentially using larger over- and under-passes. Red deer Cervus elaphus and wolves were only detected using very wide passages.

The type of passage had a greater influence on the use by fauna than the surrounding habitat variables (tree or shrub cover and the degree of human use of passages). However most of the crossing structures had


tree or shrub cover in close proximity with 83 per cent of the structures having cover less than 50 m away. Thus there was little comparison in the distance of passage from cover.

The authors highlight the necessity of mitigation effectiveness studies to analyse existing crossing structures as well as those specifically designed for fauna passage. Management recommendations including: installing different passage types (culverts, under- and over-passes) to compliment non-wildlife engineered structures; the type of passage to be installed should be focussed on the target species or faunal group in the area; the frequency of passages should depend on the size of the vertebrates that will use it (i.e. smaller species require more frequent passages than those species with larger physical and territorial sizes); and position passages in areas that are commonly used by species, thus simplifying habituation to the structure.

Foster, M.L. and Humphrey, S.R. (1995). Use of highway underpasses by Florida panthers and other wildlife. Wildlife Society Bulletin. 23 (1): 95-100.

Location of study: Interstate 75, Collier County, Florida, USA

Habitat: mosaic of forests, marshes, and disturbed areas. Forests including hardwood swamps, slash pine (Pinus elliottii) and common baldcypress (Taxodium distichum) forests or savannas. Disturbed habitats were mostly roadsides and agricultural fields.

Target species: Florida panther Felis concolor coryi

Mitigation structure/s: 4 purpose-built underpasses: 21.2 - 25.8 m wide, 48.5 m long (including 2 bridges 13.1 m and median strip 22.3 m which was open, i.e. no bridge overhead), height not given. 64 km of interstate had fence: 3 m high chain-link fence made out of galvanised steel topped with 3 strands of barbed wire.

Determined the effectiveness of fencing and underpasses in allowing movement of panthers and preventing road mortality along the Interstate 75. Four of the 24 purpose-built underpasses were monitored with digital event recorders and cameras for 2, 10, 14 and 16 months. There was also a concurrent telemetry study on radio-collared panthers and bobcats.

Cameras recorded 10 crossings by panthers, 133 by bobcats Lynx rufus, 361 by deer Odocoileus virginianus, 167 by racoons Procyon lotor, 9 by


alligators Alligator mississipiensis, and 2 by black bears Ursus americanus. Telemetry data on radio-collared Florida panthers found that the 10 crossings were by two individuals, 1 using 2 different underpasses 8 times and 1 using 1 underpass twice. Demonstrates importance of concurrent studies to investigate type of movement. Study also demonstrates the limitations of camera and tracking studies on assessing the value of structures. A few individuals repeatedly using the same underpass may affect the frequency of crossings.

Radio-collared panthers and bobcats used the underpasses mainly to move to portions of their home range separated by the highway. Numerous individual deer were photographed grazing in the underpass and racoons were photographed in pools of water. Thus the underpasses may reduce the effects of fragmentation especially for animals' that have home ranges encompassing both sides of the highway.

Time of day influenced use by different species. Panthers crossed exclusively at night. Racoons and bobcats usually crossed from dusk to dawn, deer most often crossed during the day. The lack of use in particular underpasses by some species may be due to interspecific interactions e.g. even though numerous photographs of deer were taken in three of the four underpasses, the fourth underpass may have been avoided because it was frequently used by bobcats and humans.

As a result of panthers and other animals using the underpasses, the authors infer that the underpasses reduced mortality for some species. This was not shown quantitatively. The authors suggest that the placement of underpasses along traditional wildlife trails adds to the effectiveness of the crossing structure.

Aresco, M.J. (2005). Mitigation measures to reduce highway mortality of turtles and other herpetofauna at a north Florida lake. Journal of Wildlife Management, 69: 549-560.

Location of study: US Highway 27, Lake Jackson, northwestern Florida, USA

Habitat: 1,620 ha lake that has been divided by highway,

Target species: herpetofauna

Mitigation structure/s: one drainage culvert (3.5 m in diameter, 46.6 m long) connecting two portions of lake. Constructed fence along 700 m


of Highway, designed to direct fauna into culvert. Bottom edge of fence was buried ca. 20 cm so above ground height was ca. 40 cm.

Calculated rate of mortality over a 40-day period. Built fence to direct fauna into existing culvert and evaluated the effectiveness of fence in reducing road mortality and facilitate migration over 2.5 years.

McDonald, W. and St. Clair, C.C. (2004). Elements that promote highway crossing structure use by small mammals in Banff National Park. Journal of Applied Ecology. 41:82-93.

Location of study: Trans-Canada Highway

Habitat: very brief mention as being montane habitat

Target species: three murid rodents: meadow voles Microtuspennsylvanicus, deer mice Peromyscus maniculatus, red-backed voles Clethrionomys gapperi.

Mitigation structure/s: Overpasses, underpasses and culverts: two overpasses (15 m wide and 75 m or 79 m long) covered with sparse trees and shrubs with patches of bare ground. Nine 3 m diameter arch shaped underpasses with no vegetation inside underpass or at entrance and had length between 64 and 73 m. Nine drainage culverts (0.3 m diameter and 63 to 72 m long) with introduced grasses near entrances.

Experimental study to assess the response by three murid rodent species to variation in crossing structure type, vegetation cover at the entrance, and the distance between the crossing structure and the individuals home range. Animals were trapped within 50 m of the structure entrance, moved to the other side of the highway and released adjacent to the crossing structure. Flourescent powder was applied to all animals released to monitor success and path characteristics of animals on the return home. Preference to the type of crossing structure was assessed by translocating individuals within 2 m of an overpass, underpass or culvert. To determine the role of vegetation cover at the structures entrance in attracting animals to use the structures, the amount of cover was modified at both entrances of two overpasses and nine underpasses. There were three treatments of cover: heavy (100 per cent), medium (50 per cent) or no cover (0 per cent). To quantify the distance that translocated individuals would travel to their home range via crossing structures, animals were released at 20, 40 or 60 m from underpass entrance. (Any animal that did not


return within 4 days was captured and returned manually).

54 per cent of translocated animals (n=166) returned successfully to their original capture locations and all returned via the crossing structures (except 2 animals when released 60 m from crossing structure). In the crossing structure preference experiment, successful returns were affected by crossing structure type and species. All species were more successful returning through 0.3 m diameter culvert than the other structures. These smaller structures may provide additional overhead cover making these structures seem safer from predators. Deer mice had a higher return success than the other two species through all crossing structures. The high return success of the nocturnal deer mice coincides with the lower traffic volume at night.

Crossing structures with heavy cover at the entrances elicited a greater return success rate than structures with medium or no cover provided. This suggests that the addition of cover provides an inexpensive way to increase the effectiveness of crossing structures. Heavy cover correlated with more tortuous movement paths (more complex paths), especially for meadow voles.

Meadow voles had lower return success through all structures than other species (for structure type and cover experiments) and were unable to return when no cover was provided. This reluctance in using the crossing structures may be due to their diurnal activity and the corresponding high traffic volume or by their relative lack of mobility compared with deer mice. Those individuals that were translocated 60 m from crossing structures had a very low return success. This indicates that animals are more likely to use crossing structures that were near their home range. The authors suggest that ideally crossing structures should be frequently placed which corresponds with the spatial scale that the targeted species move.

Tull, J.C. and Krausman, P.R. (2001). Use of a wildlife corridor by desert mule deer. The Southwestern Naturalist, 46: 81-86.

Location of study: water canal (Central Arizona Project), Avra Valley, Arizona, USA

Habitat: palo verde Cercidium mixed cacti on bajadas, creosote Larrea tridentata - bursage Ambrosia in undisturbed flats, mesquite Prosopis velutina - burroweed Isocoma tenuisecta in disturbed flats, ironwood


Olneya tesota - canyon ragweed Ambrosia ambrosioides in washes, and desert grassland.

Target species: desert mule deer Odocoileus hemionus eremicus

Mitigation structure/s: Tucson Mitigation Corridor (TMC), which was established to mitigate lost habitat from canal construction and facilitate movement across canal: 7 wildlife crossing areas (40 to 230 m wide) where the canal is underground and four flumes (allows water from moving downslope to pass over canal).

Investigated the use of the Tucson Mitigation Corridor (TMC) by desert mule deer. Captured deer in and around TMC in Nov 1995 and Feb 1996, and fitted with radio collars (collared 3 males and 14 females). Each radio-collared deer was found by direct observation, triangulation, helicopter or infrared camera � 16 times per season. Seasonal home ranges were estimated for each individual. From data able to identify movement patterns and determine where and when move across canal.

Four deer crossed the canal. One female's home range abutted the canal and she used the TMC to cross the canal at least three times. Two deer were observed crossing the canal outside the TMC (500 m north of canal where the canal is underground). The TMC provided access across a linear barrier (the canal), thus acting as an overall movement corridor system.

8.5 International Reports, Theses and Conference Proceedings

Furniss M J, Gubemick B& Clarkin K (2003) ‘Design and Construction of Aquatic Organsim Passage at Road-stream Crossings: Site assessment and Geomorphic Considerations in Stream Simulation Culvert’ in ICOET (2003) Proceedings- Making Connections, pp 30-52.Location of Study: Watercourses with road crossings, throughout USA

Habitat: Aquatic habitat in general

Target Species: Aquatic Species

Mitigation Structures: Culverts with stream Simulation measures

Major Linear Infrastructure: Roads and traditional culverts

This is a scientific publication that describes the types of damage road crossings with traditional culverts can cause to aquatic populations and meta-populations by limiting aquatic species movement along stream


corridors. It stipulates that road-stream crossing culverts designed in the traditional way, which are traditionally sized for some rare flood flow, have predictable detrimental effects on stream channels themselves. Considering during floods, the culverts may plug or be overtopped, and demonstrates that over time the culvert may impede downstream movement of woody debris and sediment.

This paper describes common stream responses to culverts, such as chronic aggradation and degradation; long-term changes in stream stability due to interruption of woody debris transport; and sedimentation sustained when culverts plug and fail, etc.

It also describes the range of approaches to crossing design, from a culvert sized only to pass a certain flood to valley-spanning bridges and viaducts. It then uses Stream simulation principles and places it in the context of other design approaches that provide more or less biological and geomorphic connectivity. Importantly it highlights that biological and geomorphic priorities and risks must be weighed against site constraints and costs to select the appropriate level of continuity for each site.

Finally it provides site assessment procedures for stream simulation design. These included surveying and describing the longitudinal profile and valley cross-sections, bed material assessment, and reference reach selection. Channel stability interpretations needed for design were also discussed.

Although this paper discusses and describes potential methods for applying stream simulation to individual situations it also acknowledges that this technique is still a ‘developing art/ science’ as such it recognises that this paper may not provide concrete fact or specific guidelines. It also recognises that practitioners in different geographic regions will modify it to fit each region. Hence this paper gives an appropriate outline of the potential for culverts with stream simulation in future developments but still addresses it limitations.

Furniss M J, Gubemick B& Clarkin K (2003) ‘Design and Construction of Aquatic Organsim Passage at Road-stream Crossings: Site assessment and Geomorphic Considerations in Stream Simulation Culvert’ in ICOET (2003) Proceedings- Making Connections, pp 30-52.Location of Study: Watercourses with road crossings, throughout USA


Habitat: Aquatic habitat in general.

Target Species: Aquatic Species.

Mitigation Structures: Culverts with stream simulation measures.

Major Linear Infrastructure: Roads and traditional culverts.

This is a scientific publication that describes the types of damage road crossings with traditional culverts can cause to aquatic populations and meta-populations by limiting aquatic species movement along stream corridors. It stipulates that road-stream crossing culverts designed in the traditional way, which are traditionally sized for some rare flood flow, have predictable detrimental effects on stream channels themselves. Considering during floods, the culverts may plug or be overtopped, and demonstrates that over time the culvert may impede downstream movement of woody debris and sediment.

This paper describes common stream responses to culverts, such as chronic aggradation and degradation; long-term changes in stream stability due to interruption of woody debris transport; and sedimentation sustained when culverts plug and fail, etc.

It also describes the range of approaches to crossing design, from a culvert sized only to pass a certain flood to valley-spanning bridges and viaducts. It then uses stream simulation principles and places it in the context of other design approaches that provide more or less biological and geomorphic connectivity. Importantly it highlights that biological and geomorphic priorities and risks must be weighed against site constraints and costs to select the appropriate level of continuity for each site.

Finally it provides site assessment procedures for stream simulation design. These included surveying and describing the longitudinal profile and valley cross-sections, bed material assessment, and reference reach selection. Channel stability interpretations needed for design were also discussed.

Although this paper discusses and describes potential methods for applying stream simulation to individual situations it also acknowledges that this technique is still a ‘developing art/ science’ as such it recognises that this paper may not provide concrete fact or specific guidelines. It also recognises that practitioners in different geographic regions will modify it to fit each region. Hence this paper gives an appropriate outline of the potential for culverts with stream simulation


in future developments but still addresses their limitations.

Riley C ‘Fish Passage at Selected Culverts on the Hoonah Ranger District, Tongrass National Forest in ICOET (2003) Proceedings- Making Connections, pp 73-82

Location of Study: Hoonah River, Tongrass National Forest.

Habitat: Aquatic habitat, Tongrass National Forest.

Target Species: Samonids including steelhead trout, cutthroat trout and Dolly Varden charr.

Mitigation Structures: In essence these structures are meant to be mitigation measures but prove not to be. 38 traditional culverts, 13 (anadromous) (with barrier heights ranging from 10cm to 99cm, averaging 38.8cm and gradients ranged from 0.5 percent to 9.5 percent and averaged 3.4 percent culverts) and 17 culverts in streams providing habitat cutthroat trout and Dolly Varden charr (with outlet barrier heights ranging from 0cm to 205cm, averaging 37.2cm and gradients ranging from 2.0 percent to 9.0 percent and averaged 4.3 percent).

Major Linear Infrastructure: Roads and traditional culverts.

This paper discusses a case study situation where in July 1997, 38 culverts suspected of blocking upstream passage of juvenile salmonids were investigated on the Hoonah Ranger District. It describes the attributes measured, which included

� Species/numbers of fish upstream and downstream of each culvert, and

� Physical characteristics such as outlet barrier height, culvert gradient and upstream habitat.

The results of the study revealed that thirty of the thirty eight culverts exhibited some form of physical impediment (excessive barrier height and/or gradient) to the upstream migration of juvenile salmonids. The following results were determined from the field survey of the river:

� Thirteen Class I (anadromous) culverts were sampled, of which nine lacked juvenile cohort upstream of the culvert (no juvenile steelhead trout were trapped during the study). Boxplots of number of juvenile salmon trapped upstream of culverts related a considerable reduction in distribution, median, and mean as compared to downstream. All 13 culverts exhibited an outlet


perched above the streambed, with barrier heights ranging from 10cm to 99cm, averaging 38.8cm.

� Nineteen culverts were identified as Class II culverts (i.e., culverts in streams providing cutthroat trout and Dolly Varden charr habitat occupied upstream of anadromous habitat) during the survey, of which 17 exhibited some physical form of barrier to juvenile passage. Eight of the 19 Class II culverts had resident fish species trapped upstream of the culvert, six of which occurred above culverts exhibiting barrier characteristics such as outlet perch or excessive gradient.

� Boxplot distribution, median, and mean of height of outlet barrier and culvert gradient tended to be greater at Class I and Class II structures without fish trapped upstream as compared to culverts where fish were trapped upstream. This pattern was repeated for culvert length at Class I crossings, but was reversed for Class II structures.

� Overall, barrier culverts resulted in a loss of 8.11km (16,534m2) of fish habitat, comprised of 2.68km (6,408m2) of Class I and 5.42km (10,126m2) Class II habitat.

In essence this study demonstrates how traditional culverts designed for flood plans have a significant impact on the passage of fish, particularly less aggressive fish, to upstream habitat, when looking at the overall loss of habitat across a large portion of a river. Although this paper admits that describing habitat lost using only length measures simplifies discussion of this subject, the primary focus of topic is not lost.

The scientific methods and analysis utilised for this study appear to be sound considering sample size was sufficient to produce representative and random samples. As such a basic assessment of the study would suggest that the scientific rigour of this paper is creditable.

8.6 International Reviews and Syntheses

Little, S.J., Harcourt, R.G. & Clevenger, A.P. (2002) Do wildlife passages act as prey-traps? Biological Conservation 107: 135-145

Reviews available information on crossing structures to address the


question of whether they act as ‘prey-traps’. That is, whether a higher concentration of animals at wildlife passages is exploited by predators. The information surveyed included published scientific literature and proceedings of the International Conference on Wildlife Ecology and Transportation (1996, 1998 and 1999). Some opinion was also sought from researchers in the field.

The following topics were addressed:

� Use of passage structures by predators and prey

A variety of predators have been found using wildlife passages regularly, for movement or as part of their territories. They may also use them for shade and shelter. Many potential prey species are also known to use these structures. This includes seasonal migration and dispersal, or daily movement.

� Do predators use artificial structures to facilitate hunting success?

Carnivores vary in their response to roads and similar infrastructure, with some species favoured and others having reduced populations in such areas. Some predators will exploit artificial corridors to increase their foraging opportunities, and roads can provide greater access to prey. Mitigation structures have been used to improve hunting success, e.g., chasing and trapping prey against exclusion fencing. There have also been reports of tunnels with a high level of predator use coinciding with low use by prey species. This does suggest that predators will take advantage of artificial structures to increase their foraging opportunities, but does not provide evidence they are facilitating prey capture.

� Direct and indirect evidence of hunting in and near passages

Predators have been reported hunting in or near wildlife passages. There is no evidence of increased predator density or predation rates at passages. No study has examined predator-induced prey mortality at passage sites compared with control areas away from passages. There have also been no before-after studies of the impact of passage construction on predator-prey interactions. Further research is required in this area.

� Do prey species avoid passages frequented by predators?

Observational studies show that prey species will use the same passages as predators. Coincidental use may be masking actual patterns in community structure. Some studies do suggest predator


species may use different passages to prey, but the reasons for this are unclear. It is difficult to separate the impact of predator-prey interactions from habitat differences and passage structural attributes on species use. This needs further investigation.

� Do prey species use passages at different times than predators?

Few studies have looked at this. No study has examined whether the timing of use of passages by predator and prey species varies from normal periods of activity or if predator-prey avoidance behaviour is being exhibited. Only two studies were found of passage use in relation to predator and prey daily activity patterns. In Florida, panthers used passages only at night while racoons and bobcats usually crossed from dusk to dawn and white-tailed deer most often crossed during the day. In New South Wales, most fox and native mammal activity was nocturnal, while cat and dog activity occurred both day and night. The reviewers hypothesise that timing avoidance may be more likely in areas where predators and prey evolved together, but less likely when introduced species are involved.

� Mechanisms for predator avoidance at passage structures

Prey species have morphological and behavioural adaptations against predation. Behavioural defences may involve predator avoidance or escape. As wildlife passages are generally more exposed and restricted/narrow, it may limit the effectiveness of these behaviours. Cover at entrances may provide some protection, but also gives some predators the opportunity to ambush prey. Early detection of predators by prey (e.g., using scent) may be more important at wildlife passages. This may be more likely when the species have co-evolved. The reviewers hypothesise that structures are more likely to act as prey-traps in areas where the major predators are introduced, e.g., in Australia.

� The influence of human activity

Studies of the influence of human activity on the use of passages by predator and prey species have given mixed results. Humans may reduce species use by hunting or creating disturbance. Human activity had an important influence on use in some cases. For example, in Banff large carnivores were most influenced by human activity, while structural and landscape attributes had more of an influence on use of passages by ungulates. The influence of human disturbance on passage use has not been studied in Australia. The reviewers suggest that where


predator and prey have co-evolved passage use by predators may be more susceptible to human activity. It is therefore possible that in Australia the influence of human activity would be less marked, and may be positively related to predators, as they are often associated with human habitation and disturbance.

Overall, it appears that there is no evidence of passages commonly being exploited as prey-traps. However, no studies have specifically examined predation rates in or near passages compared to areas further away.

Implications for passage design are discussed. It is suggested that a range of passages in different habitat types may be preferable to relying on one structure for all species. This may also assist prey species to better detect or avoid predators. For example, short, open passages give larger prey species such as deer more chance to detect predators. In Australia, it is thought that vegetation cover at passage entrances makes them more attractive to native prey species and attracts less feral predators. Predator-proof fencing may be appropriate in some cases, but this may also exclude larger native species. Baiting predators at passage sites is also an option, though the impact on non-target species needs to be considered.

Suggestions for future research are also discussed.

van der Ree R., van der Grift E. A., Mata C. & Suarez F. (2007) Overcoming the barrier effect of roads – how effective are mitigation strategies? An international review of the effectiveness of underpasses and overpasses designed to increase the permeability of roads for wildlife. In: International Conference on Ecology and Transportation (eds C. L. Irwin, D. Nelson and K. P. McDermott) pp. 423-31. Center for Transportation and The Environment, North Carolina State University, Raleigh, NC, Little Rock, Arkansas, USA.

This review paper summarises 123 studies on the use of a range of wildlife crossing structures published in consultants reports, theses, conference proceedings and refereed journal articles. It included literature published in English, French, German, Spanish and Dutch. An important distinction was made to differentiate between use of a structure and its effectiveness at achieving a certain goal. This paper summarises a range of content from each study, including geographic


location, number and type of wildlife crossing structure studied, the type of linear infrastructure, study design, duration and timing of study, the extent to which the populations and habitat adjacent to the road was described, survey technique, and whether the factors influencing rate of use was identified.

Of the 123 studies reviewed, all but two recorded use of some of the crossing-structures by wildlife. Two studies demonstrated an effect at the population level, indicating that the extent to which population viability has increased subsequent to installing wildlife crossing structures remains unclear. A total of 1864 wildlife crossing structures were reported on, with the majority (83 per cent) underpasses, and specifically culverts (40 per cent of 1864). Other types of structures included tunnels, bridges, land bridges, canopy bridges, and glider poles. The majority of structures traversed roads, and it was rare to find studies that fully detailed the dimensions or conditions of either the crossing structure or linear infrastructure. Most studies commenced after the mitigations structures were installed, and a before-after approach was evident in just 15 of the 123 studies. The duration of monitoring ranged from 4 days to 20 years. Approximately half of the studies included some assessment of the size of populations in habitat adjacent to the structures. The most common form of animal survey was by identifying tracks in a suitable medium, followed by the use of cameras. While most studies included an assessment of factors influencing rate of use (98 of 123), only 24 used a quantitative approach.

8.7 International ‘Best Practice’ Manuals and Guidelines

Vereniging Das and Boom: Association for the preservation of mustilidae and their habitat in The Netherlands (1990). Provisions for badgers against traffic.

Location of study: n/a information for public and managers concerning badgers and road traffic

Habitat: n/a

Target species: badgers

Mitigation structure/s: guidelines for tunnel, fence and one-way


swing gate construction

This paper provides information for the general public and managers on how to minimise badger road-kill. The first tunnel built in Holland was in the mid 1970's during the construction of new motorways. At the time of publication, some 50 tunnels existed with agreements for the construction of a further 60 by 1995. The paper discusses the most effective dimensions of tunnels and fences with recommendations on the best construction materials to use. There is also an emphasis on erecting badger proof fences on both sides of the road and placing tunnels at regular intervals. Also mentioned is the importance of ensuring that the route/s to tunnels are safe and are more attractive to badgers by planting a variety of plants.


RE

FE

RE

NC

ES

9 References

Abson R. (2004) The use by vertebrate fauna of the Slaty Creek wildlife underpass, Calder Freeway, Black Forest, Macedon, Victoria. University of Tasmania.

Abson R. and Lawrence R. E. (2003a) Monitoring the use of the Slaty Creek wildlife underpass, Calder Freeway, Black Forest, Macedon, Victoria, Australia. In 'Proceedings of the 2003 International Conference on Ecology and Transportation'. Lake Placid, New York, USA. (Eds CL Irwin, P Garret and KP McDermott) pp. 303-308. (Center for Transportation and Environment, North Carolina State University)

Abson R. and Lawrence R. E. (2003b) 'Slaty Creek wildlife underpass study, Final Report.' (Centre for sustainable regional communities, Latrobe University, Bendigo: Kyneton)

Amor R. L. and Stevens P. L. (1976) Spread of weeds from a roadside into sclerophyll forest at Dartmouth, Australia. Weed Research 16, 111-118.

Angold P. G. (1997) The Impact of a Road Upon Adjacent Heathland Vegetation - Effects On Plant Species Composition. Journal of Applied Ecology 34, 409-417.

Australian Museum Business Services (2001a) 'Fauna underpass monitoring Stage One - Final Report - Brunswick Heads. Roads and Traffic Authority.' Australian Museum Business Services, Sydney, NSW.

Australian Museum Business Services (2001b) 'Fauna underpass monitoring Stage One - Final Report - Buladelah to Coolongolook. Roads and Traffic Authority.' Australian Museum Business Services, Sydney, NSW.

Australian Museum Business Services (2001c) 'Fauna underpass monitoring Stage One - Final Report - Herons Creek. Roads and Traffic Authority.' Australian Museum Business Services, Sydney, NSW.

Australian Museum Business Services (2001d) 'Fauna underpass monitoring Stage Two, Episode Three, Bulahdelah to Coolongolook.' Australian Museum Business Services, Sydney.


Australian Museum Business Services (2001e) 'An investigation of the use of road overpass structures by arboreal marsupials.' Australian Museum Business Services, Sydney, Sydney.

Australian Museum Business Services (2001f) 'Pacific Highway - Fauna underpass monitoring Stage Two, Episode Three Taree.' Australian Museum Business Services, Sydney, NSW.

Australian Museum Business Services (2001g) 'Pacific Highway - Fauna underpass monitoring Stage Two, Episode Three Taree.' Australian Museum Business Services, Sydney, NSW.

Australian Museum Business Services (2002a) 'Fauna underpass monitoring Stage Two, Episode Five Buladelah to Coolongolook.' Australian Museum Business Services, East Sydney, NSW.

Australian Museum Business Services (2002b) 'Fauna underpass monitoring Stage Two, Episode Five, Taree.' Australian Museum Business Services, East Sydney, N.S.W.

Baker J., Goldingay R. and Whelan R. J. (1998) Powerline easements through forests: a case study of impacts on avifauna. Pacific Conservation Biology 4, 79 - 89.

Ball T. & Goldingay R. (2008) Can wooden poles be used to reconnect habitat for a gliding mammal? Landscape and Urban Planning 87, 140-6.

Barnett J. L., How R. A. and Humphries W. F. (1978) The use of habitat components by small mammals in eastern Australia. Australian Journal of Ecology 3, 277-285.

Baur A. and Baur B. (1990) Are roads barriers to dispersal in the land snail Arianta arbustorum? Canadian Journal of Zoology 68, 613-617.

Bennett A. F. (1990) 'Habitat Corridors. Their Role in Wildlife Management and Conservation.' (Department of Conservation and Environment: Melbourne)


Bennett A. F. (1999) 'Linkages in the Landscape. The Role of Corridors and Connectivity in Wildlife Conservation.' (IUCN: Gland, Switzerland and Cambridge)

Bond A. and Jones D. N. (2008) Temporal trends in use of fauna-friendly underpasses and overpasses. Wildlife Research 35, 103-112

Braden A. W., Lopez R. R., Roberts C. W., Silvy N. J., Owen C. B. and Frank P. A. (in press) Florida Key deer underpass use and movements along a highway corridor. Wildlife Biology.

Brown G. P., Phillips B. L., Webb J. K. and Shine R. (2006) Toad on the road: Use of roads as dispersal corridors by cane toads (Bufo marinus) at an invasion front in tropical Australia. Biological Conservation 133, 88-94.

Brown R. J., Brown M. N. and Pesotto B. (1986) Birds killed on secondary roads in Western Australia. Corella 10, 118-122.

Cain A. T., Tuovila V. R., Hewitt D. G. and Tewes M. E. (2003) Effects of a highway and mitigation projects on bobcats in Southern Texas. Biological Conservation 114, 189-197.

Clarke D. J., Pearce K. A. and White J. G. (2006) Powerline corridors: degraded ecosystems or wildlife havens? Wildlife Research 33, 615-626.

Clark W. D. and Karr J. R. (1979) Effects of highways on red-winged blackbird and horned lark populations. Wilson Bulletin 91, 143-145.

Clevenger A. P., Chruszcz B. and Gunson K. (2001) Drainage culverts as habitat linkages and factors affecting passage by mammals. Journal of Applied Ecology 38, 1340-1349.

Clevenger A. P., Chruszcz B. and Gunson K. E. (2003) Spatial patterns and factors influencing small vertebrate fauna road-kill aggregations. Biological Conservation 109, 15-26.

Clevenger A. P. and Waltho N. (1999) Dry drainage culvert use and design considerations for small- and medium-sized mammal movement across a major transportation corridor. In 'Proceedings of the third


international conference on wildlife ecology and transportation'. Missoula, Montana. (Eds G Evink, P Garrett and AD Ziegler). (Florida Department of Transportation)

Clevenger A. P. and Waltho N. (2000) Factors influencing the effectiveness of wildlife underpasses in Banff National Park, Alberta, Canada. Conservation Biology 14, 47-56.

Clevenger A. P. and Waltho N. (2003) Long-term, year-round monitoring of wildlife crossing structures and the importance of temporal and spatial variability in performance studies. In 'Proceedings of the 2003 International Conference on Ecology and Transportation'. Lake Placid, New York, USA. (Eds CL Irwin, P Garrett and KP McDermott) pp. 293-302. (Center for Transportation and Environment, North Carolina State University)

Clevenger A. P. and Waltho N. (2005) Performance indices to identify attributes of highway crossing structures facilitating movement of large mammals. Biological Conservation 121, 453-464.

Coulson G. (1989) The effect of drought on road mortality of macropods. Australian Wildlife Research 16, 79-83.

Cuperus R., Canters K. J., Udo de Haes H. A. and Friedman D. S. (1999) Guidelines for ecological compensation associated with highways. Biological Conservation 90, 41-51.

Davenport J. and Davenport J. L. (2006) (Ed.)^(Eds) 'The ecology of transportation: managing mobility for the environment.' Environmental Pollution (Springer: Dordrecht, The Netherlands)

Davies J. M., Roper T. J. and Sheperdson D. J. (1987) Seasonal distribution of road kills in the European badger (Meles meles). Journal of Zoology 211, 525-529.

Develey P. F. and Stouffer P. C. (2001) Effects of roads on movements by understory birds in mixed- species flocks in central Amazonian Brazil. Conservation Biology 15, 1416-1422.

Dhindsa M. S., Sandhu J. S., Sandhu P. S. and Toor H. S. (1988) Roadside birds in Punjab (India): in relation to mortality from vehicles. Environmental Conservation 15, 303-310.


Dodd C. K., Barichivich W. J. and Smith L. L. (2004) Effectiveness of a barrier wall and culverts in reducing wildlife mortality on a heavily traveled highway in Florida. Biological Conservation 118, 619-631.

Donaldson A. and Bennett A. F. (2004) 'Ecological effects of roads. Implications for the internal fragmentation of Australian Parks and reserves.' (Parks Victoria: Melbourne)

Ecologia Environmental Consultants (1995) 'Kwinana Freeway wildlife underpass study.' Ecologia Environmental Consultants, Mount Lawley, W.A.

Fahrig L., Pedlar J. H., Pope S. E., Taylor P. D. and Wegner J. F. (1995) Effect of road traffic on amphibian density. Biological Conservation 73, 177-182.

Fitzgerald M. (2003) 'Results of fauna monitoring in designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: February - March 2003.' Mullumbimby, NSW.

Fitzgerald M. (2004) 'Results of sandtrap monitoring in 14 designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: Sample 2 October 2003 - January 2005.' Mullumbimby, NSW.

Fitzgerald M. (2005) 'Final Report. Results of sandtrap monitoring in 8 designated fauna crossings of the Yelgun to Chinderah Pacific Highway Upgrade: Sample 3 February - April 2005.'

Forman R. T., Reineking B. and Hersperger A. M. (2002a) Road traffic and nearby grassland bird patterns in a suburbanizing landscape. Environmental Management 29, 782-800.

Forman R. T. T. and Deblinger R. D. (2000) The ecological road-effect zone of a Massachusetts (USA) suburban highway. Conservation Biology 14, 36-46.

Forman R. T. T., Sperling D., Bissonette J. A., Clevenger A. P., Cutshall C. D., Dale V. H., Fahrig L., France R., Goldman C. R., Heanue K., Jones J. A., Swanson F. J., Turrentine T. and Winter T. C. (2002b) 'Road Ecology. Science and Solutions.' (Island Press: Washington)


Foster M. L. and Humphrey S. R. (1995) Use of highway underpasses by Florida panthers and other wildlife. Wildlife Society Bulletin 23, 95-100.

Gibbons P. and Lindenmayer D. B. (2007) Offsets for land clearing: No net loss or the tail wagging the dog? Ecological Management & Restoration 8, 26-31.

Gibbs J. P. (1998) Amphibian movements in response to forest edges, roads, and streambeds in southern New England. The Journal of Wildlife Management 62, 584.

Gloyne C. C. and Clevenger A. P. (2001) Cougar Puma concolor use of wildlife crossing structures on the Trans-Canada highway in Banff National Park, Alberta. Wildlife Biology 7, 117-124.

Goldingay R. L. and Whelan R. J. (1997) Powerline easements - do they promote edge effects in eucalypt forest for small mammals? WildlifeResearch, 737-744.

Goosem M. (1997) Internal fragmentation: the effects of roads, highways and powerline clearings on movements and mortality of rainforest vertebrates. In 'Tropical forest remnants: ecology, management and conservation of fragmented communities'. (Eds WF Laurance and RO Bierregaard jr.) pp. 241-255. (University of Chicago Press: Chicago)

Goosem M. (2001) Effects of rainforest roads on small mammals: inhibition of crossing movements. Wildlife Research 28, 351-364.

Goosem M. (2004) Linear infrastructure in the tropical rainforests of far north Queensland: mitigating impacts on fauna of roads and powerline clearings. In 'Conservation of Australia's Forest Fauna (Second edition) '. (Ed. D Lunney) pp. 418-434. (Royal Zoological Society of New South Wales: Mosman, NSW)

Goosem M. (2005) Effectiveness of rope bridge arboreal overpasses and faunal underpasses in providing connectivity for rainforest fauna. In 'Proceedings of the 2005 International Conference on Ecology and Transportation'. Raleigh, North Calrolina, USA. (Eds CL Irwin, P Garrett and KP McDermott) pp. 304-316. (Center for Transportation and the Environment, North Carolina State University)


Goosem M. and Marsh H. (1997) Fragmentation of a small mammal community by a powerline corridor through tropical rainforest. Wildlife Research 24, 613-629.

Guyot G. and Clobert J. (1997) Conservation measures for a population of hermanns tortoise Testudo hermanni in southern France bisected by major highway. Biological Conservation 79, 251-256.

Hamer A. O., A (2006) 'Craigieburn Bypass: Monitoring of crossing structures for the growling grass frog Litoria raniformis, Thomastown to Craigieburn, Victoria. Final Report prepared for VicRoads.' Ecology Partners Pty Ltd.

Harris I. (2007) Use of wildlife underpasses by quenda (Isoodonobesulus fusciventer) at Roe Highway Stage 7. Honours Thesis, Faculty of Natural and Agricultural Sciences, The University of Western Australia.

Haxton T. (2000) Road mortality of Snapping Turtles, Chelydraserpentina, in Central Ontario during their nesting period. Canadian Field-Naturalist 114, 106-110.

Hayes I. (2006) Effectiveness of fauna road-kill mitigation structures in north-eastern New South Wales. Southern Cross University.

Hels T. and Buchwald E. (2001) The effect of road kills on amphibian populations. Biological Conservation 99, 331-340.

Hunt A., Dickens H. J. and Whelan R. J. (1987) Movement of mammals through tunnels under railway lines. Australian Zoologist 24, 89-93.

Irwin C. L., Garrett P. and McDermott K. P. (2003) (Eds). '2003 Proceedings of the International Conference on Ecology and Transportation.' (Centre for Transportation and Environment, North Carolina State University: Raleigh, North Carolina)

Irwin C. L., Garrett P. and McDermott K. P. (2005) (Eds). 'Proceedings of the 2005 International Conference on Ecology and Transportation.' (Center for Transportation and Environment, North Carolina State University: Raleigh, North Carolina)


Jones D., Appleby R., Edgar J. and Green B. (2004) 'Compton Road wildlife movement program. A report for Environment and Parks, Brisbane City Council.' Suburban Wildlife Research Group, Australian School of Environmental Studies, Griffith University, Nathan, Qld.

Jones M. E. (2000) Road upgrade, road mortality and remedial measures: impacts on a population of eastern quolls and Tasmanian devils. Wildlife Research 27, 289-296.

Koenig J., Shine R. and Shea G. (2001) The ecology of an Australian reptile icon: how do blue-tongued lizards (Tiliqua scincoides) survive in suburbia? Wildlife Research 28, 215.

Lamont B. B., Rees R. G., Witkowski E. T. F. and Whitten V. A. (1994a) Comparative size, fecundity and ecophysiology of roadside plants of Banksia hookeriana. Journal of Applied Ecology 31, 137-144.

Lamont B. B., Whitten V. A., Witkowski E. T. F., Rees R. G. and Enright N. J. (1994b) Regional and local (road verge) effects on size and fecundity in Banksia menziesii. Australian Journal of Ecology 19, 197-205.

Lee E. (2006) The ecological effects of sealed roads in arid ecosystems. University of New South Wales.

Little S. J., Harcourt R. G. and Clevenger A. P. (2002) Do wildlife passages act as prey-traps? Biological Conservation 107, 135-145.

Lonsdale W. M. and Lane A. M. (1994) Tourist vehicles as vectors of weed seeds in Kakadu-National-Park, northern Australia. Biological Conservation 69, 277-283.

Mader H. J. (1984) Animal habitat isolation by roads and agricultural fields. Biological Conservation 29, 81-96.

Mansergh I. M. and Scotts D. J. (1989) Habitat continuity and social organisation of the mountain pygmy-possum restored by tunnel. Journal of Wildlife Management 53, 701-707.


May S. A. and Norton T. W. (1996) Influence of fragmentation and disturbance on the potential impact of feral predators on native fauna in Australian forest ecosystems. Wildlife Research 23, 387-400.

McDonald W. and St Clair C. C. (2004) Elements that promote highway crossing structure use by small mammals in Banff National Park. Journalof Applied Ecology 41, 82-93.

Ng S. J., Dole J. W., Sauvajot R. M., Riley S. P. D. and Valone T. J. (2004) Use of highway undercrossings by wildlife in southern California. Biological Conservation 115, 499-507.

O'Donnell J. (2003) Practical and Cost Effective Terrestrial Fauna Friendly Design on the Pacific Highway. Powerpoint Presentation.

Osawa R. (1989) Road-kills of the swamp wallaby, Wallabia bicolor, on North Stradbroke Island, south-east Queensland. Australian Wildlife Research 16, 95-104.

Pedevillano C. and Wright R. G. (1987) The influence of visitors on mountain goat activities in Glacier National Park, Montana. Biological Conservation 39, 1-11.

Pohlman C. L., Turton S. M. and Goosem M. (2007) Edge effects of linear canopy openings on tropical rain forest understory microclimate. Biotropica 39, 62-71.

Port G. R. and Thompson J. R. (1980) Outbreaks of insect herbivores on plants along motorways in the United Kingdom. Journal of Applied Ecology 17, 649-656.

Ramp D. and Ben-Ami D. (2006) The effects of road-based fatalties on the viability of a peri-urban Swamp Wallaby population. Journal of Wildlife Managament 70, 1615-1624.

Reed D. F. (1981) Mule deer behavior at a highway underpass exit. Journal of Wildlife Management 45, 542-543.

Reed D. F., Woodard T. N. and Pojar T. M. (1975) Behavioral response of mule deer to a highway underpass. Journal of Wildlife Management 39, 361-367.


Reh W. and Seiz A. (1990) The influence of land use on the genetic structure of populations of the common frog Rana temporaria. Biological Conservation 54, 239-249.

Robertson P. (2002) Discussion Paper. Design requirement for structures to ameliorate the potential effects on frog movements of construction and operation of the proposed Craigieburn Bypass Freeway. p. 23. Wildlife Profiles, Heidelberg, Victoria.

Rodriguez A., Crema G. and Delibes M. (1997) Factors affecting crossing of red foxes and wildcats through non-wildlife passages across a high-speed railway. Ecography 20, 287-294.

Roedenbeck I. A., Fahrig L., Findlay C. S., Houlahan J. E., Jaeger J. A. G., Klar N., Kramer-Schadt S. and Van der Grift E. A. (2007) The Rauischholzhausen agenda for road ecology. Ecology and Society 12, 11. [online] URL: http://www.ecologyandsociety.org/vol12/iss1/art11/.

Rosen P. C. and Lowe C. H. (1994) Highway mortality of snakes in the Sonoran desert of southern Arizona. Biological Conservation 68, 143-148.

Scott L. K., Hume I. D. and Dickman C. R. (1999) Ecology and population biology of long-nosed bandicoots (Perameles nasuta) at North Head, Sydney Harbour National Park. Wildlife Research 26, 805-821.

Scotts D. J. & Seebeck J. (1989) Ecology of Potorous longipes (Marsupialia: Potoroidae); and preliminary recommendations for management of its habitat in Victoria. In: Arthur Rylah Institute for Environmental Research Technical Report Series No. 62. Department of Conservation, Forests and Lands, Heidelberg, Victoria.

Seabrook W. A. and Dettmann E. B. (1996) Roads as activity corridors for cane toads in Australia. Journal of Wildlife Management 60, 363-368.

Shuttleworth C. (2001) Traffic related mortality in a red squirrel (Sciurus vulgaris) population receiving supplemental feeding. Urban


Ecosystems 5, 109-118.

Singer F. J. (1978) Behavior of mountain goats in relation to US Highway 2, Glacier National Park, Montana. Journal of Wildlife Management 42, 591-597.

Singer F. J. and Doherty J. L. (1985) Managing mountain goats at a highway crossing. Wildlife Society Bulletin 13, 469-477.

Spellerberg I. F. (1998) Ecological effects of roads and traffic: a literature review. Global Ecology & Biogeography Letters 7, 317-333.

Spooner P. G. (2005) Response of Acacia species to disturbance by roadworks in roadside environments in southern New South Wales, Australia. Biological Conservation 122, 231-242.

Statham M. and Statham H. L. (1997) Movements and habits of brushtail possums (Trichosurus vulpecula Kerr) in an urban area. Wildlife Research 24, 715-726.

Swihart R. K. and Slade N. A. (1984) Road crossing in Sigmodon bispidus and Microtus ochrogaster. Journal of Mammalogy 65, 357-360.

Taylor B. D. and Goldingay R. L. (2003) Cutting the carnage: wildlife usage of road culverts in north-eastern New South Wales. WildlifeResearch 30, 529-537.

Valladares Padua C., Cullen Jr L. and Padua S. (1995) A pole bridge to avoid primate road kills. Neotropical Primates 3, 13-15.

van der Ree R. (2006) Road upgrade in Victoria a filter to the movement of the endangered Squirrel Glider Petaurus norfolcensis: Results of a pilot study. Ecological Management and Restoration 7, 226-228.

van der Ree R., Soanes K., Cesarini S., McCall S., Taylor A., Sunnucks P., Harper M., Gulle, N. (2008). Rope bridge story revisited. Episode 30 of The University of Melbourne 'Visions' Video podcasts. www.visions.unimelb.edu.au, accessed 12 August 2008.


van der Ree R., McCarthy M. A., Heinze D. and Mansergh I. M. (in review) Under-road tunnel restores population dynamics of endangered pygmy-possum. Ecology and Society.

Vestjens W. J. M. (1973) Wildlife mortality on a road in New South Wales. Emu 73, 107-112.

Vos C. C. and Chardon J. P. (1998) Effects of habitat fragmentation and road density on the distribution pattern of the moor frog Rana arvalis. Journal of Applied Ecology 35, 44-56.

Wasser S. K., Bevis K., King G. and Hanson E. (1997) Non-invasive physiological measures of disturbance in the northern spotted owl. Conservation Biology 11, 1019-1022.

Yanes M., Velasco J. and Suarez F. (1995) Permeability of roads and railways to vertebrates: the importance of culverts. Biological Conservation 71, 217-222.


AP

PE

ND

IX 1

Summary of Criteria Against which each Publication was Assessed

Criterion Description / Explanation

Study details 9.1.1

Authors Author of paper or report

Title Tile of paper or report

Abbreviated title 9.1.1

Publication type SCI = scientific, refereed journal, BOO = book, CON = conference proceedings, THE = thesis, REP = consultants report, REV = review,

Date of publication 9.1.1

Publication details Name of book, journal title, volume, page numbers etc

Language ENG = english, NL = Dutch, ESP = spanish, GER = German

Region which area the study was conducted in

State/Province etc. State or province study conducted in

Country AUSTralia, USA, CANada, SPAIN, FRANCE, NETHerlands

Study design 9.1.1

Hypothesis stated YES/NO/UNCLEAR. This gives an indication of what the authors expected to find

Hypothesis give details

Aims stated YES/NO

Aims give details

Study design BACI, BA, CI, CORrelative study, REGression approach, STAT = descriptive stats only, NOSTAT = no data given on rate of use or number of crossings etc

Controls used? YES/NO

Control details give details

What was measured CROSS = crossing rate, KILL = road-kill, BEH = animal behaviour, DISP = dispersal, SURV = survival, USE = use of structure, PRESENCE, GENE flow, HOME range,

Level of study INDIV = individuals, single species, INDIV2 = individuals of multiple species, POP = population of a single species, POP2 = populations, multiple species

Date of study 9.1.1

Duration of study number of weeks, months years

Frequency of monitoring

9.1.1

Measure of success given

This acts to define an endpoint in mitigation enhancement. E.g. when road-kill = 20 individuals per year, or when gene flow = 40% per year, then we will stop. How do we know when the structure is good enough?

Measure of success give details of how they measured success


Crossing away from structure measured

YES/NO

Crossing away from structure

give details of method

Problem clearly defined YES/NO/UNCLEAR

Details of problem e.g. road mortality of xx%, genetic separation of two populations

Extrapolation to or study of population level

9.1.1

Statistical analyses 9.1.1

Roads/traffic 9.1.1

Description of road given

YES/NO/PART (PART is where road is partly described - e.g. road lanes but not width)

Description of road Give details

Description of traffic given

YES/NO/PART (PART is where traffic is partly described - e.g. traffic volume at certain time of day)

Description of traffic give details

Type of linear infrastructure

Road, rail

Number of roads/linear infrastructure

Number of roads, railway lines studied

Length of roads Over what length of road were the crossing structures / mitigation positioned?

Structure 9.1.1

Type of structure Tunnel, culvert, overpass, underpass, glider pole, canopy bridge, deerguard

Size of structure given YES/NO/PART (part when some of Length, width, height not given)

Size of structure give details

Number of structures total number of structures monitored (not the number that may have been built)

Number of culverts number of culverts studied

Number of tunnels number of tunnels studied

Number of underpasses

number of underpasses studied

Number of overpasses number of overpasses studied

Intended purpose Intended purpose of structures - WILD = purpose built for wildlife, WATER = drainage, WATERMOD = originally drainage, but later modified for use by wildlife

Target species 9.1.1

Cost given Cost to build structure YES/NO

Cost Give details

Habitat 9.1.1

Vegetation description YES/NO/PART


given

Vegetation give brief details

Altitude, geology, topography etc. given

YES/NO/PART

Altitude, geology, topography

give details

Wildlife 9.1.1

Focus (e.g, endangered, ferals, common species)

Endangered or rare species, feral animals, native vertebrates, locally significant species, safety = animal status not given, but focus on human safety/road-kill

Assessed presence/abundance in area

YES/NO

Method to detect wildlife in area

Trap = live trapping, tracks (in sand, snow, soot, inkpads etc), radiotracking, count = game counter, camera = remotely triggered camera, obs = direct observations, gen = genetic data, scat = scat analysis, hair = hair traps, spotlighting, frog calls, bird calls, bat = electronic bat detectors

Assumed presence YES/NO

Details of assumption of presence

give details = e.g. anecdotal records, museum records, likely suitable habitat

Method for measuring use

TRAP = live trapping, TRACKS (in sand, snow, soot, inkpads etc), RADIOtracking, COUNT = game counter, CAMERA = remotely triggered camera, OBS = direct observations, GEN = genetic data, SCAT = scat analysis, HAIR = hair traps, SPOTlighting, FROG calls, BIRD calls, BAT = electronic bat detectors

Presence in structure OBS = observed or EXP = expected based on a measure of local abundance

Species detected give details

Through-passage identified

YES/NO (if yes - may not be perfect, but indicates an attempt to determine complete crossing)

Method for determining through-passage

give detail

Type of use If identified - e.g. DISPERSAL, MIGRATION, DAILY (e.g. foraging, home range activity), EXPLORE

Most common species using structure

give details

Factors identified as influencing movement?

If yes - give details

Fencing 9.1.1

Fencing used YES/NO/UNCLEAR

Fencing details FUNNELLING, FENCED (to keep animals off road), UNFENCED, OTHFUNNEL (other things utilised to funnel wildlife e.g. rocks)

Notes 9.1.1


Miscellaneous Engineering Drawings

Following are some drawing showing:-

Drawing Set A: Shows two typical arch under passes (designated 1A & 1B); and a typical box culvert.

Drawing Set B: Shows designs in progress of various fauna fences, including: round and box culverts with fauna fencing; fauna and frog fences; and ‘fauna escapes’ and fauna fencing.

Drawing Set C: Shows various glider crossing drawings including glider crossing detail sheet 1,

Drawing Set D: Shows long and short underpasses with fauna rails.

CT001.1108