Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=uhdw20

Human Dimensions of WildlifeAn International Journal

ISSN: 1087-1209 (Print) 1533-158X (Online) Journal homepage: http://www.tandfonline.com/loi/uhdw20

Using Participatory Methods to Assess Data PoorMigrant Fisheries in Kenya

Innocent Ngao Wanyonyi, Denis Macharia, Adel Heenan & Stephen C. Mangi

To cite this article: Innocent Ngao Wanyonyi, Denis Macharia, Adel Heenan & Stephen C. Mangi(2018): Using Participatory Methods to Assess Data Poor Migrant Fisheries in Kenya, HumanDimensions of Wildlife, DOI: 10.1080/10871209.2018.1488304

To link to this article: https://doi.org/10.1080/10871209.2018.1488304

Published with license by Taylor & FrancisGroup, LLC © 2018 Innocent NgaoWanyonyi, Denis Macharia, Adel Heenan,and Stephen C. Mangi.

Published online: 26 Jun 2018.

Submit your article to this journal

View Crossmark data

Using Participatory Methods to Assess Data Poor MigrantFisheries in KenyaInnocent Ngao Wanyonyi a,b, Denis Machariab,c, Adel Heenand, and Stephen C. Mangie

aDepartment of Biological and Environmental Sciences, School of Natural Sciences, Linnaeus University,Kalmar, Sweden; bSocMon, Coastal Oceans Research and Development in the Indian Ocean (CORDIO),Mombasa, Kenya; cTechnical Services Directorate, Regional Centre for Mapping of Resources for Development(RCMRD), Nairobi, Kenya; dSchool of Ocean Sciences, Bangor University, Anglesey, UK; eCentre forEnvironment, Fisheries and Aquaculture Science (CEFAS), Menai Bridge, Plymouth, UK

ABSTRACTSpatial information is limited for artisanal fisheries management andalmost entirely absent for migrant fishers. Here, we addressed thisdata gap for East African migrant fishers via participatory mappingmethods. We worked with 14 migrant fishing vessels operating fromfour fish landing sites in Kenya. We monitored individual vesselsusing GPS tracking to produce fishing ground intensity maps. Wethen generated fishing preference maps via focus group discussions.The fishing intensity maps provided high-resolution spatial informa-tion on fishing activities, whereas the fishing preference maps iden-tified preferred fishing grounds. These two techniques generallyshowed high agreement. By further integrating these two fishercoproduced maps with supplemental vessel logbook data, it is clearthat any spatial management measures would most affect migrantfishers using ringnets, hook and line, and cast nets gear. Our success-ful application of low-technology participatory mapping techniquesto provide geospatial fisheries data have broad application to datapoor fisheries worldwide.

KeywordsArtisanal fisheries; fishingpreference; GPS tracking;mapping; PGIS

Introduction

Artisanal fisheries are typically data poor. People participating in these fisheries usuallyoperate on a small-scale, via low-technology methods, using small and/or traditionalfishing vessels and gear, and are limited in range. In Kenya, for example, artisanal fisherstend to fish adjacent to their home landing site at inshore shallow reefs and lagoons,fishing for local consumption, and sale (Obura & Wanyonyi, 2001). Even less informationis available on migrant fishers, which is a type of artisanal fisher that travels to fish awayfrom their home fishing areas for periods ranging from a few weeks to few months(Wanyonyi, Wamukota, Tuda, Mwakha, & Nguti, 2016). Migrant fishers are flexible intheir fishing practices and often operate in remote locations less accessible to fisheriesmanagement authorities (Binet, Failler, & Thorpe, 2012; Islam & Herbeck, 2013; Rosendo,Jiddawi, Joubert, Mechisso, & Brown, 2008). Consequently, fishing activities of migrantartisanal fishers, as in the case of Kenya, are particularly difficult to monitor and manage.

CONTACT Innocent Ngao Wanyonyi innocent_ke@yahoo.com Barlastgatan 11, SE - 392 31, Kalmar, Sweden.Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uhdw.

HUMAN DIMENSIONS OF WILDLIFEhttps://doi.org/10.1080/10871209.2018.1488304

Published with license by Taylor & Francis Group, LLC © 2018 Innocent Ngao Wanyonyi, Denis Macharia, Adel Heenan, and Stephen C. Mangi.

This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License(http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium,provided the original work is properly cited, and is not altered, transformed, or built upon in any way.

Migrant fishers in East Africa, mainly from Pemba and Zanzibar, arrive in large numbersalong the Kenyan coast and are a common component of small-scale fisheries of this region.The importance of these fishing grounds to migrant fishers is evident as they often equal oroutnumber local fishers at many fish landing sites (Wanyonyi, Wamukota, Mesaki,Guissamulo, & Ochiewo, 2016). The sources and destinations, as well migration patternsof migrant fishers are well documented (Fulanda et al., 2009; Wanyonyi et al., 2016), but thelocation of their main fishing grounds are not well understood (Scholz, Steinback, Kruse,Mertens, & Silverman, 2011). The lack of understanding the spatial extent and fishingpatterns of migrant fishers prevents their integration into local resource managementprocesses (Campbell, Stehfest, Votier, & Hall-Spencer, 2014). There is an urgent need tofill this information gap due to the fast growing numbers of artisanal fishers in the region(Ochiewo, 2004).

Typically, artisanalfisheries are assessed via catch estimates from fish landings sites, whereasthe intensity of resource extraction is assessed separately through ecological surveys that areseldom spatially explicit or linked to specific fishing grounds (McClanahan, Glaesel, Rubens, &Kiambo, 1997; McClanahan &Mangi, 2004). In industrial fisheries, the spatial distribution offishing activity can be determined using surveillance data from observer vessels and electroniclogbook catch data (e.g., Bastardie, Nielsen, Ulrich, Egekvist, & Degel, 2010; Forcada, Valle,Sánchez-Lizaso, Bayle-Sempere, & Corsi, 2010; St. Martin, 2006). In artisanal fisheries, how-ever, small vessel size and the small-scale nature of the fisheries operations limit the applicationof such techniques. Data collection, therefore, employs methodologies based on self-sampling,logbooks, interviews, and direct observation (Cinner, McClanahan, & Wamukota, 2010;Kimani & Obura, 2007; Lokrantz, Nyström, Norström, Folke, & Cinner, 2009; Malleret-King et al., 2003; Mangi, Smith, & Catchpole, 2016). Spatial data collection in these fisherieshas focused on the use of participatory methods such as sketch maps and diagrams drawn byfishers to document their spatial access and fishery resource uses (Kimani & Obura, 2007;Obura & Wanyonyi, 2001). The application of real time tracking to generate high-resolutionhuman related spatial data in artisanal fisheries ismore recent. Studies have assessed the fishingbehavior of small-scale fishers following innovations such as the introduction of fish-aggregat-ing devices to the fishery (Alvard, McGaffey, & Carlson, 2015).

The application of precise and cost-effective tools for objectively quantifying spatial aspects inartisanal fisheries remain limited (Cassels, Curran, & Kramer, 2005). In particular, the applica-tion of real time tracking technology has not yet been rolled-out in most artisanal fisheries. Ageographic information system (GIS) offers a potential solution for collecting spatially explicitdata on artisanalfisheries. GIS is defined as “a computing application capable of creating, storing,manipulating, visualizing, and analysing geographic information” (Dunn, 2007, pp. 616–617). AGIS consists of a complex interconnection of three essential components: computer hardware,GIS software, and trained human personnel that can be applied in a particular situation. It is notcheap to setup a GIS, as a modern computer capable of running this system costs about US$2,100, the ArcGIS software is US$1,700, and personnel require specialized GIS training to runthe analysis and produce useful final map outputs. In addition, a GIS requires large and accuratedatasets to run. The application of GIS technology in many other spheres of developing countrysettings faces similar issues and challenges (Sipe&Dale, 2003). The application ofGIS to artisanalfisheries, particularly the collection of spatially explicit data on human related aspects of thefisheries, is also a challenge.

2 I. N. WANYONYI ET AL.

One potential solution is through participatory GIS (PGIS). PGIS is a data collectionmethod that gives greater privilege and legitimacy to local or indigenous spatial knowledgeand emphasizes community involvement in the production and/or use of geographicalinformation (Dunn, 2007, pp. 616–617). PGIS involves the resource users in research (inthis case, fishers), enabling the prioritization of the collection of information that is mostrelevant to them. It includes intermediate outputs in the form of maps that can be imme-diately useful and the final outputs are processed from a centralized GIS, limiting the costs.

Study Objectives

The main goal of our multi-method study was to characterize the migrant fishing fleetoperating off the Kenyan coast. Specifically, our objectives were to: (a) document the spatialfishing patterns of migrant fishers using fishing preference mapping and GPS tracking, (b)compare fishing preference maps generated by fishers with the GPS-tracking data, and (c)integrate the GPS-tracking data with fishing vessel logbook data to generate scores ofimportance for fishing grounds per landing site. We achieve this via an interdisciplinaryapproach to demonstrate the utility of PGIS to collect and map high-resolution and reliableinformation on fishing behavior and target resources of artisanal fishers. By following migrantfishers at four sites in Kenya to identify the fishing areas they frequently use and getting fisherstomap their preferred fishing grounds, we present a novel way of gatheringmuch needed datathat could be applied to migratory fishers elsewhere around the world.

Methods

Study Location

Nearshore coastal fisheries in Kenya are governed at the local level by a beach managementauthority (BMU). The BMU controls access to fishing grounds and landings via the local fishlanding site. Migrant fishers seek permission from this authority to fish in the local community’swaters. Typically, permission is sought upon arrival into an area to fish these waters and to landcatch at the particular landing site for the duration of their migration (Wanyonyi et al., 2016).Migrant fishers who arrive at the destination together and belong to the samefishing crewwouldreturn home together, usually after 1–7 months (Wanyonyi et al., 2016). The main destinationlanding sites and places of origin of migrant fishers in East Africa have been identified (Fulandaet al., 2009;Wanyonyi et al., 2016). Our study focuses on four of the most common destinationsfor migrant fishers originating from Pemba (Figure 1).

The study sites compromised of the Kipini, Gazi, and Shimoni villages, and theassociated fish landing site and fishing grounds. The villages of Vanga and Jimbo weregrouped and treated as one additional site (called Vanga-Jimbo here) due to the small sizeof Jimbo. These two villages are in close proximity to each other and are ethnicallyhomogenous, so we deemed this grouping to be appropriate.

Data Collection

Data were collected from October 2010 to March 2011, to coincide with the annual arrivalof migrant fishers in the largest numbers (Wanyonyi et al., 2016). We used two main

HUMAN DIMENSIONS OF WILDLIFE 3

participatory methods to collect data and map fishing activities. We obtained informationon fishing activities using self-reported information via fishing vessel logbooks to supple-ment the GPS-tracking data.

Mapping Fishing Activity via GPS TrackingAt the start of the migrant season, six migrant fisher vessels were randomly selected ateach study site. Vessel crew members were invited to be trained in how to use andhandle the GPS units (Garmin etrex). We then selected three of the vessels at each site(except at Gazi where two vessels were selected due to availability of GPS units) toparticipate in the study for the remaining duration of their migrant trip. Vessels wereselected on the basis of being representative of the main fishing vessel-gear types usedby migrant fishers operating from each village landing site. A crew member on each ofthe monitored vessels was designated as a data collector. At the onset of each fishingexcursion, the data collector switched on the GPS unit and switched it off upon returnto the landing site. The GPS units were programmed to mark location every 2 min,thus tracking the location of the vessel throughout the fishing trip. The GPS trackswere retrieved from the data collector every 8–10 days and downloaded via EasyGPSsoftware (TopoGrafix, 2016). Our method was adopted from Van der Spek, VanSchaick, De Bois, and De Haan (2009).

At the start of the study, 11 vessels were equipped with the GPS-tracking units.During the study, three vessels moved away from their original landing site (Shimoni,Gazi, and Vanga-Jimbo). We replaced these with alternative vessels. In addition, onevessel departed the area with the GPS unit leading to a loss of eight tracking days inShimoni in the analysis. In total, we tracked 14 crews during the fisher migrationseason. After accounting for these losses, the total number of tracking days was 80, 83,73, and 98 in Kipini, Gazi, Shimoni, and Vanga-Jimbo, respectively. Here, we present

Figure 1. Map of study locations.

4 I. N. WANYONYI ET AL.

the GPS-tracking maps and fishing preference maps for the vessels with the longesttracking duration per site. Prior to collecting data on the location of individual fishingareas, we obtained permission from the vessel captains and the officials fromeach BMU.

Mapping Fishing Activity via Area Preference MapsThe second participatory mapping component of the study involved mapping coastal andmarine resources at each study site (i.e., fish landing sites, beaches, fishing grounds, reef areas).We conducted 11 focus group discussions with 2–5 migrant fishers representing eachparticipating vessel. Fishers annotated a previously prepared and geo-referenced base map,adding details on the community’s fishing grounds with names and main reefs, channels, andbeaches. They identified different landmarks on the map including the coastline, fish landingsites, roads, villages, and rivers. The information wasmapped byGPS and digitized to produceparticipatory resource maps for each study site (e.g., Aswani & Lauer, 2006; Léopold,Guillemot, Rocklin, & Chen, 2014). The fishing areas were marked on the maps andcategorized into main, least fished, and reserved areas (e.g., Furletti, Cintia, Renso, &Spinsanti, 2013). Reserve fishing areas refer to areas where respective fisher’s crew left freefrom fishing for a certain period. Fishers identified reserve areas including fish spawningaggregation sites and deep channels that are only fished under certain circumstances such asduring the Northeast Monsoon season (Pers. Comm., migrant fisher, Kipini).

Characterizing Fishing Activity via Fishers Log BooksCrewmembers of the GPS tracked vessels were trained to record their daily fishing tripinformation using a logbook. This included noting the following information for eachfishing trip: catch size (kg) and the main composition (kind) and value of catch landed.Logbook information was merged with vessel details including number of fishers onboard, vessel, and fishing gear type, and the physical characteristics of vessels (length,engine size, main gear, fish landing site). All activities by the vessel during fishing werecaptured including zero catch dates and non-fishing days (e.g., Bastardie et al., 2010).Results, including mean and standard errors, are presented for all vessels.

Data Analysis

GPS TrackingAs GPS devices are usually continuously recording the vessel position as calibrated, theunprocessed track log data were checked for validity and integrity, and filtered. Each positionwas tested for polygon inclusion and all inland positions except the last position before leavingthe fish landing site were removed before analysis. Each consecutive position outside the fishlanding site, termed here as the point of commencement, constitutes part of a fishing trip (i.e.,from point of departure to arrival back at the fish landing site). Valid trajectories wereconverted into point shape files, using Esri ArcCatalog, and imported into Esri ArcGIS version10. Layers containing trajectories for each GPS unit at each fish landing site were spatiallymerged into single point layers. Fishnets made up of square polygons measuring 1 km by 1 kmdraping over the resource base for each site were created as a basis for analysis. Each squarepolygon had a unique identifier code. We used a ‘points to polygon’ analysis approach tocalculate the number of points (i.e., density) in each square polygon.

HUMAN DIMENSIONS OF WILDLIFE 5

During a fishing trip, the vessel exhibits sequences of stops and moves. When the vesselspends more time per unit area, this appears as high intensity on the map, whereas travelmovements are faster and appear as low intensity. High intensity suggests fishing activity at apoint of interest (POI) (e.g., Furletti et al., 2013; Van der Spek et al., 2009; Wall, Wittemyer,Klinkenberg, LeMay, & Douglas-Hamilton, 2013). To determine the amount of time spent byeach fishing boat at a particular location (represented by the square polygons), we multipliedthe number of points by two. This correction value was used since the GPS was programmedto mark points at a frequency of 2 min. Later, we converted the polygons into raster layers at1 km spatial resolution and applied a majority filter to create final raster surfaces. The totalarea for actual resource bases for each vessel in each site was obtained by digitizing polygonsaround pixels that had greater than zero, and we used Esri ArcGIS 10 to estimate total areas insquare kilometers (km2).

Fishing Area PreferenceFollowing Yates and Schoeman (2013), we used spatial access priority mapping (SAPM) toget fishers to document their fishing ground preferences in a manner that can be used forcreating quantitative maps. The advantage of SAPM is that it relies only on simpleformulas available in standard spreadsheets and GIS software to capture fisher preferences.Specifically, in the focus group discussions, the groups of migrant fishers categorized theirfishing grounds into ‘main,’ ‘least,’ and ‘reserve’ fishing grounds by assigning themimportance values of 80%, 15%, and 5%, respectively.

We calculated spatial access priority (SAP) for vessels that produced their fishingpreference maps (Table 3). The SAP is a weighted measure of how important a fishingarea is, relative to the number of fishers who operate on vessels and the size of the area.The SAP was determined for each fishing area by multiplying the number of full-timecrew on the vessel by the area’s importance value, and then dividing by the area covered(number of square kilometers):

SAP ¼ Crew� Importance=Area km2ð Þ

We identified the types of vessels and gear used by migrant fishers at each fish landingsite, and estimated average vessel lengths using information from the BMU, local leaders,and observation. There were 133 active migrant fisher vessels, which were disaggregatedaccording to fish landing site, gear type, and vessel type and length. We then calculated theaverage crew for each category (vessel type – vessel length – gear type), yielding 12categories (hereafter referred to as vessel-length-gear) (Figure 3). This number of cate-gories is higher than available types of vessels and gear combinations because the size ofcrew varies depending on the type of gear used even where the vessels are the same in size.

We employed a simple approach by making tables using basic cell formulae andcalculations in Excel. We estimated crew size for each vessel-length-gear category for allactive migrant vessels and the average crew per vessel. We created a final weightingscheme for each length-gear combination, by dividing the estimated number of crewwith the actual number sampled:

Weighting ¼ Estimated crew=sampled crew

6 I. N. WANYONYI ET AL.

This weighting was used to scale-up the sampled vessels to make estimates for thefishery. Priorities for each area were multiplied by their corresponding weights to obtain afinal weighted priority score per fish landing site by fishery (Figure 3). Weighting by thenumber of crew assigns higher total priority for vessels with more crew. This methodhighlights the relevance of size of a vessel and type, gear type, and fish landing site ininfluencing access to fishing grounds.

Fisher LogsWe assigned a common identifier for each GPS trip to a logbook trip so that the GPStracks and logbook trip data could be integrated. The logbook maintained by each vesselserved to supplement information from the GPS tracking. Logbook data were coded andentered into spreadsheets. Summary data were then generated for each vessel by aggregat-ing and analyzing trip level metrics on the main gear type in use, the catch composition(kg) identified to the family level, and the average daily income per vessel and per fisheron each vessel.

Results

Migrant Fishing Fleet Characteristics

In total, 14 migrant fishing vessels operating from four landing sites were tracked duringthe migrant fishing season of 2010–2011. These vessels included 126 individual crewmembers, approximately 20% of the total number of migrant fishers operating across allsites during the study. The main fishing vessel types at the four landing sites were largedugout, outrigger canoes, and wooden plank boats (Table 1). Main gear types weresharknet, hook and line, long line, and ringnets, as well as cast nets, drift nets, scoopnets, harpoon, and basket traps. The vessel in Shimoni covered the shortest averagedistance (15 km) to fishing grounds, whereas that of Kipini covered the longest distance(28 km) and utilized the largest fishing area (271.8 km). Shimoni had deepest averagedepth of fishing grounds (420 m) compared to other sites including Vanga-Jimbo (77.5 m)(Table 1).

Analysis of logbook data provided further detail on characteristics of the migrant fleetand individual fisher returns. Specifically, wooden plank boats, operating with scoop netsand harpoons with an average of 18–20 fishers per day, had the highest average daily catch

Table 1. Estimated number of migrant fisher vessels at landing sites during the period, October 2010 toMarch 2011, and description of fishing areas for 14 participating vessels.

Estimated number of vessels Fishing areas description (averages)

Fish landingsite

Dugoutcanoes

Outriggercanoes

Wooden plankboats

Sport fishingboats

Distance(m)

Depth(m)

Fishing area(km2)

Kipini 10 1 7 0 28 111.7 271.8Gazi 6 9 4 1 21.3 277.5 139.6Shimoni 65 8 2 0 15 420 149.4Vanga-Jimbo 9 4 7 0 17 77.5 118.4Total 90 22 20 1Max 40 510 455.4Min 9 10 43.5Ave 20.1 215.4 164.0

HUMAN DIMENSIONS OF WILDLIFE 7

(72.4 kg), whereas the dugout canoes using shark nets with three fishers per day had thelowest (17.6 kg) (Table 2). The foreign exchange rate at the time of this study was US$1for KSh 85. Dugout canoes targeting sharks had the lowest vessel daily income (KSh1,224.21), whereas the dugout canoes operating cast nests with 17–18 fishers per day hadthe highest (KSh 5,557.18). Individually, fishers working on outrigger canoes using hookand line gear with three fishers per day had the highest daily income (KSh 600.11),whereas those fishing on wooden plank boats using scoop nets and harpoons had thelowest (KSh 249.10). The difference in individual versus vessel level incomes is attributableto the higher number of fishers working on the wooden plank boats (Table 2). Asexpected, given the gear types in use, the wooden plank boats and dugout canoes usingscoop nets, harpoons, and shark nets, respectively, were more specific in their target catch(namely octopus and sharks and rays). In contrast, dugout canoes using cast nets andoutrigger canoes using hook and line were less selective, targeting more than four maintypes of catch (Table 2).

Spatial Fishing Patterns and Comparison of Mapping Methods

The GPS-tracking data identified areas of high intensity use by wooden plank boatsoperating from Kipini as clusters of patch reef adjacent to Ziwayuu Island and twonortherly reefs in shallow water (<10 m deep) habitat (Figure 2A). From the logbookdata, we know that these vessels use scoop net gear to fish for octopus and lobsters.

Fishing area preference mapping identified Mwamba Mwakamanga, Mwamba Punju,and Mwamba Ziwayuu as preferred fishing grounds, corresponding to areas of high-fishing intensity from GPS tracking (Figure 2). The reserve-fishing area at Ras Tenewicorresponded to no fishing activity on the GPS-tracking map. However, Mwamba Mateweidentified as least fished, but was a POI with intensive use according to GPS tracking.

GPS tracking showed the POIs for dugout canoes using a cast net in Gazi were spread outalong the steady slopes of the fringing reef adjacent to Gazi Bay, southward to Funzi Island(Figure 2B). Fishing areas were mostly in shallow habitat in waters of 10m depth. The areas ofhigh intensity identified by the GPS-tracking data corresponded to main fishing areasidentified by fisher preference maps at Kwale and Mpungani (Figure 2B). As expected, fishersdid not use reserve sites identified on the preference map. However, an intensive use area offMsambweni was marked on the preference map as least fished at Arusha and Mekka.

Table 2. Aggregated catch and income for selected migrant vessels.

Fish landing siteVesseltype Gear type

Main catch composition(kg)

Averagedaily

catch (kg)

Average dailyincome pervessel (KSh)

Average dailyincome perfisher (KSh)

Kipini Woodenplankboat

Scoop netandharpoon

Octopus 72.4 4,802.00 240.10

Gazi Dugoutcanoe

Cast net Mackerel, Half beak,Sardines, Barracuda

51.1 5,557.18 308.73

Shimoni Outriggercanoe

Hook andline

Emperor, Groupers, Jacksand Trevallies, Parrot fish,Snappers

38.5 1,800.33 600.11

Vanga-Jimbo Dugoutcanoe

Shark net Rays, Sharks 17.6 1,224.21 408.07

Note. At the time of study 1US$ = 85 KSh.

8 I. N. WANYONYI ET AL.

POIs for outrigger canoe fishers using hook and line in Shimoni were clustered within asmall area of patch reef within shallow areas up to 20 m deep (Figure 2C). These fishersidentified Nyuli as most fished in their preference mapping, and an offshore area over400 m deep as reserve-fishing grounds. Both areas on fishing preference maps corre-sponded to POI and no fishing activity, respectively, on their GPS map. However,Mpunguti ya chini, Ngushini, and Langoni identified as least fished were intensive useareas according to GPS tracking.

For the dugout canoes using shark nets in Vanga, intensive use areas were concentratedoff Funzi Island, on the boundary of Kisite-Mpunguti marine protected area (MPA), andat Sii Island (Figure 2D). Fishing areas in the north were up to 50 m deep and less than10 m in the south. However, GPS tracking showed high-fishing intensity at areas identifiedas reserve-fishing grounds. These included Funzi Island, Mijira, and to the southernboundary of the Kisite-Mpunguti MPA. Sii Island identified as least fished was shownto be an intensive use area by GPS maps.

Figure 2. GPS tracking (i) and Fishing Suitability (ii) maps for migrant fishers (A) at Kipini using woodenplank boat with scoop net and harpoon, (B) at Gazi using dugout canoe and Cast net.

HUMAN DIMENSIONS OF WILDLIFE 9

Table 3. Estimated fishing priority (SAP) for active vessels at four fish landing sites.Migrant fishery characteristics Calculated SAP, km2 at respective landing sites

Vessel type Gear type Kipini Gazi Shimoni Vanga-Jimbo

Dugout canoe Basket traps 3.3 40.0 4.0Dugout canoe Drift net 4.0Dugout canoe Long line 4.3Dugout canoe Shark net 5.4 14.4Dugout canoe Cast net 485.2Outrigger canoe Drift net 5.1 2.6 1.3Outrigger canoe Hook and line 34.0 169.8 203.8 101.9Wooden plank boat Hook and line 6.0 2.0 4.0Wooden plank boat Long line 2.5Wooden plank boat Shark net 3.6 5.4Wooden plank boat Ringnet 295.2 316.3Wooden plank boat Scoop net harpoon 70.1 0.0Total 117.7 963.0 572.5 130.9

Figure 3. GPS tracking (i) and Fishing Suitability (ii) maps for migrant fishers (A) at Shimoni usingoutrigger canoe and Hook and Line and (B) at Vanga-Jimbo using dugout canoe and shark net.

10 I. N. WANYONYI ET AL.

Priority Fishing Grounds for Each Landing Site

Gear-vessel combinations of ringnets-wooden plank boats, cast net-dugout canoes, and hookand line-outrigger canoes had larger SAP than dugout canoes and wooden plank boats usingbasket traps and long lines, respectively (Table 3). In general, Gazi had the highest estimated SAPfollowed by ringnets at Shimoni andGazi, and hook and line at Shimoni (Figure 3). Gazi had thehighest total combined SAP followed by Shimoni, whereas Vanga-Jimbo and Kipini had theleast. Total resource areas fromGPS tracking and SAPs for cast net, hook and line, and shark netwere comparable, but varied remarkably for long line, scoop-harpoon, and basket trap (Figure 4).

Discussion

Determining how migrant fishers use space, their fishing operations, and overlap withlocal community fisher activities are critical for their integration into fisheries manage-ment. Collecting information on this notoriously data poor component of the artisanalfishery system is challenging when the capacity and resources available to do so areseverely limited. Without this information, the impact of migrant fishers on the widersocio-ecological system will remain unknown (Cassels et al., 2005; Curran, 2002). Here, weadopted a multi-method, participatory approach using GPS tracking, fishing area pre-ference mapping, and fisher logbooks to gain insight into this often neglected componentof fishery systems. By doing so, we demonstrated the utility of low-tech solutions tounderstand migrant fisher behavior. Our study aim was to characterize migrant fishing by:(a) documenting their spatial fishing patterns, (b) identifying fishing ground preferences

Figure 4. Migrant fishers’ Suitability Area Priority (SAP) from fishing preference mapping and ResourceArea from GPS tracking by type of gear, vessel and landing site.

HUMAN DIMENSIONS OF WILDLIFE 11

and the level of catch from different fishing grounds, and (c) generating scores ofimportance for fishing grounds for particular fishers and landing sites.

Spatial Assessment of Migrant Fishers and the Complementarity of PGIS Tools

The areas of intense fishing activity as identified by GPS tracking and the fishing areasshown by the preference mapping exercise were generally in high agreement. In addition,these two methods yielded complementary types of information. GPS tracking providesfine-scale spatial and seasonal mapping of fishing activity and spatial allocation of fishingeffort, whereas preference mapping identifies the important fishing areas to fishers andfacilitates analysis of access priorities by different fisheries. The spatial data on fishermovement was further supplemented by logbook data, which was used for estimatingspatially explicit catch landings values and fishing efforts. There is a clear spatial differ-entiation of fishing activity that is determined by migrant fisher gear and vessel type.

Unsurprisingly, some mismatches in the categorization of fishing grounds by GPStracking and fishing suitability mapping were apparent. These discrepancies betweenactual fished areas from GPS tracking and self-reported main fishing grounds fromsuitability mapping do not necessarily indicate unreliability of self-reported informationby fishers. For instance, a regular fishing route may appear as high-intensity use, whereas areserve-fishing ground may fall under high-intensity use due to a difference in interpreta-tion of terms used by the fisher and scientist. Reserve-fishing grounds are areas aparticular fisher’s crew left free from fishing for certain periods and they can fish theseareas only at specific times. These include known fish spawning areas and areas of strongcurrents (Pers. com, migrant fisher, Kipini). Fisher knowledge of important fish aggrega-tion and spawning areas and their awareness of the community conservation areas (CCAs)may contribute to unwillingness to discuss such areas if they spend more fishing effort inthese areas. The fishing mapping preference approach has the advantage of documentingthe fishing ‘seascape’ that is unique to the migrant fishing community from the perspectiveof its members (Silvano & Begossi, 2012; St. Martin, 2001). This information, which isessential to fisheries governance, has until now been obscured by the assumption thatcommunity members, such as fish traders and local community fishers, represent theinterests of the entire fishing community (Nielson et al., 2004).

Using Low-Tech PGIS Methods to Engage Fishers in Data Sparse Fisheries

Migrant fishers are typically excluded from fisheries management consultations. Both GPStracking and fishing area preference mapping can engage otherwise underrepresented orignored migrant fishers. However, preference mapping using suitability mapping allowsfishers to consciously contribute knowledge-based quantitative information, whereas GPStracking uses direct automated mapping aided by a GPS unit. The SAPM approachemploys a simple method suitable for engaging resource users in fisheries research. Itdocuments fishers’ actual access priorities to provide insight into their most importantparts of the sea, based on their expertise in fishing areas. As a result, SAPM is a usefultechnique to map resource use patterns when GPS tracking may be prohibitively costly(due to the need for GPS units).

12 I. N. WANYONYI ET AL.

This cost-effective PGIS approach also has the advantage of harnessing fisherknowledge, which because it is aggregated overtime, can provide a better understand-ing of local fisheries resource use patterns (Cassels et al., 2005). Experienced fisherspossess cumulative and detailed spatial knowledge about fishing areas and theirenvironment (St. Martin, 2001). They rely on this personal knowledge to determinetheir fishing success, often repeatedly visiting choice fishing locations with higherprobability of encountering their target fish (Pet-Soede, Van Densen, Hiddink, Kuyl,& Machiels, 2001). Migrant fishers use such knowledge to target rich fishing groundswith available fish stocks, which are likely to have large concentrations of fish or areoff-limits to most local fishers, including fishing in close proximity to conservationareas, coralline islands, and offshore reefs. Migrant fishers, therefore, can provideunique information on the offshore locations where they operate. PGIS can capturethis knowledge gained by fishers during their daily interactions with their environment(Cassels et al., 2005; St. Martin & Hall-Arber, 2008). PGIS documents fisher placenames for key fishing grounds, revealing otherwise unknown seascapes (Silvano &Begossi, 2012; St. Martin, 2001). Without the engagement of migrant fishers, theseremote parts of the fishery system would otherwise be largely absent from fisheriesassessments and management (Binet et al., 2012), and knowledge of the local geogra-phy of the fishery system would remain limited to local fishers who mainly operateinshore and close to their home fish landing sites (Obura & Wanyonyi, 2001).

Relevance of Participatory Approaches in Artisanal Fisheries Management

In Kenya, BMUs are responsible for recording landings and enforcing fisheries regulations alongthe coastline and ‘inshore waters’ (Government of Kenya, 2007, 2012). This responsibilityincludes receiving migrant fishers at their respective landing sites (Wanyonyi et al., 2016). Thelimited spatial jurisdiction of a BMU affects their ability to manage migrant fishers becausemigrant fishing grounds can extend beyond the inshore waters and into adjacent areas that areunder the jurisdiction of neighboring BMUs. As a result, local fisheries management initiativesare based on ad hoc decision-making without data on the fishing operations of migrant fishers.This omission has resulted in the varied acceptance of migrant fishers, stifling gains made inartisanal fisheries management (Wanyonyi et al., 2016).

The sharing of fishing grounds between migrant fishers and fishers from neighboring land-ings sites are met with mixed tolerance of migrant fishers. Mkunguni, Mwaembe, andMwandamo landing sites were among the BMUs in Kwale County that did not allow migrantfishers. These BMUs share fishing grounds with Gazi, Shimoni, and Vanga-Jimbo landing sites,which are important destinations for migrant fishers. The presence of migrant fishers, therefore,accentuates possibilities of resource related conflicts. Indeed, increased migrant fisher activitieswithin the traditional fishing zones used by local communities have previously resulted in violentconflicts (Binet et al., 2012; Glaesel, 1997; Ochiewo, 2004).

The overlap in migrant fishing grounds between BMUs that do and do not accept migrantfishes that we identified is particularly relevant to the increasing regional interest in fisheriescomanagement including CCAs. CCAs are a form of MPA led by the local community(McClanahan, Muthiga, & Abunge, 2016). The spatial characterization that we have providedof the migrant fisher communities working in the area, which includes their landing sites andfishing gear used, can help anticipate future conflict and inform collaborative planning for future

HUMAN DIMENSIONS OF WILDLIFE 13

CCAs. Furthermore, we are able to identify: (a) important fishing areas for fishers targetingdifferent target resources and (b) themost vulnerable fishers who stand to lose the largest fishingarea due the implementation of a conservation area. As a result, we find this PGIS approach to bea viable planning or zoning tool for supporting selection or creation of local protected area siteswith minimal potential conflicts among stakeholders.

Novelty of this Approach

Artisanal fisheries research has traditionally collected data using surveys, interviews, activity logsmaintained by fishers, ecological surveys, and catch monitoring at fish landing sites (Samoilys,Osuka, Maina, & Obura, 2017). However, these methods are not spatially explicit, nor do theydifferentiate between migrant and local fishers. A spatially explicit assessment of fisher resourceuse patterns requires high resolution and reliable information about spatial and temporalpatterns and allocation of fishing effort for informed decision-making (Bastardie et al., 2010;Campbell et al., 2014; Pet-Soede et al., 2001). In addition, research methods typically employedfor studying artisanal fisheries are often criticized for distortions related to recall, reliability,reproducibility, behavioral change, and privacy issues, especially when used in isolation(Gonzalez, Hidalgo, & Barabasi, 2008; Malleret-King, Glass, Wanyonyi, & Pomeroy, 2006;Ron, Muhando, & Francis, 1998; Stone & Shiffman, 2002).

We addressed these concerns through the novel application of GPS tracking and areasuitability mapping methods to collect spatial information on migrant fisher operations. Thesemethods have been widely used in wildlife tracking (Rodgers, 2001; Wall et al., 2013), partici-patory forest ecology, and tourism usemapping (Brown, Schebella, &Weber, 2014), and here wesuccessfully integrated them via a PGIS approach. Table 4 summarizes the key features of thethree PGIS methods discussed in this article. By combining these three methods, we takeadvantage of the strengths associated with each method individually. For improved accuracyof the information gathered, we suggest these methods should be used together to complement,confirm, and validate the inferences made.

Challenges, Limitations, and Recommendations for Future Application

A major challenge to the adoption of new technologies in resource-poor settings is cost. GPStracking is a relatively cheap alternative to Vessel Monitoring Systems, which are typicallyapplied to commercial fisheries. The total project cost was less than US$910 per landing site.Core costs included threeGPS units (US$250 each), 12 rechargeable batteries (4 for each vessel atUS$35), and a shared battery charger per landing site at US$20. We limited costs by usingrechargeable batteries and asking participating fishers to activate their units at the start of fishingand switch them off on arrival back to the fish landing site to save battery power. Batteries canincrease expenditures given that dry alkaline cells have limited battery life.

An additional cost that should be considered when assessing the suitability of our approachfor use elsewhere is the time and resources required for training the participating fishers forquality assurance and harmonization of data collection. This cost links directly to the risk thatfishers could abandon the study at any time.We replaced four participating vessels after they leftthe study area (leading to a loss of data and equipment). The uncertainty and unpredictability ofmigrant fisher operations (Wanyonyi et al., 2016), will generally remain outside of the influenceof the research or implementation team. A viable alternate that could overcome this challenge is

14 I. N. WANYONYI ET AL.

the use of automatic identification system (AIS) technology. AIS has only recently becomeavailable for artisanal fishing vessels use (e.g., Proud, Browning, & Kocak, 2016). The advantagesof AIS over GPS tracking are that the device can be installed on any boat regardless of size and ituses an autonomous tracking system for two-way exchange of navigational informationalbetween vessels or AIS-equipped stations along the coast within range of VHF radio, normallyup to about 20 nautical miles (nm). AIS provides the boat’s identity and GPS data (location,course, speed) (Gezelius, 2007). AIS can not only be applied for the participatory mapping offishing grounds, but also for mapping of sensitive ecosystems such as spawning areas forconservation purposes, and navigational hazards can also be mapped for safety. AIS can, there-fore, produce additional benefits and is more efficient for recording fishing operations. These

Table 4. Key features of the three PGIS methods applied in the artisanal fisheries research.

Feature

PGIS methods

Preference mapping GPS tracking Fisher logbooks

Engagement of stakeholdersin research

Fishers engaged Fishers engaged Fishers engaged

Mode of knowledgegeneration

Fishers consciouslycontribute knowledge-basedquantitative informationbased on their experience

Trained fishers collectinformation by directautomated mapping offishing grounds aided by aGPS unit

Self-reported informationby trained fishers,collecting daily fishingoperation data linked toGPS tracking

Cost effectiveness A cost-effective approachsuitable for involvingdiverse, otherwiseunrepresented stakeholders

Relatively cheap, butincludes the cost ofacquisition of GPS units andtraining in GIS, that may behigh for some small-scalefisheries

Cost-effective in linkingfishing operations andcharacteristics to GPS dataand preference mapping

Level of technology and easeof use

No-tech PGIS using simplemethod, suitable forengaging resource users inresearch, but requires GISsoftware and skills todevelop final fishingpreference maps

Low-tech PGIS, relativelysimple method for involvingresource users in research,but requires regular trainingof fishers in GPS use. Alsorequires GIS software andskills to develop finalintensity maps.

No-tech method, relativelysimple, but requirestraining of fishers in recordkeeping. May require GISsoftware and skills foradvanced analysis andlinkages with GPS trackingmaps and preference maps

Output Documents fishers’ mostimportant fishing areas,providing a map of keyfishing grounds. It allows foranalysis of actual accesspriorities by differentfisheries and landing sites

Identifies fishing areasfrequently used by fishers byproviding fine-scale spatialmapping of their fishingactivity, thereby allowing foranalysis of spatial allocationof fishing effort

Documents fishing tripinformation including catchsize (kg), main composition(kind), and value of catchlanded, and allows forestimation of catch value,the value of the landings,and fishing effort linked tofishing ground andlocation

Allows for documentation offisher place names for keyfishing grounds, includingplace names to revealfishers’ unknown seascapes

Documents characteristicsof the fishing operation interms of number of fisherson board, vessel andfishing gear type, vessels(length, engine size, maingear, fish landing site), andallows for their analysis inrelation to fishingpreference maps togenerate suitability of useanalysis

HUMAN DIMENSIONS OF WILDLIFE 15

wider benefits are more likely to incentivize fisher participation in studies relative to datacollected purely for research.

Conclusion

We have spatially assessed the resource use and fishing patterns of migrant fishers by mappingthe location and boundaries of their target fish grounds and activities. This information fills acritical gap for artisanal fisheriesmanagement in Kenya that is particularly useful for future CCAplanning. Our study advances migrant fisher research through the application of PGIS tools anddemonstrates the viability of low-cost and low-technology approaches for collecting data ontropical fisheries and integrating migrant fishers into management decision-making processes.

Acknowledgments

The authors sincerely thank the migrant and local fishers involved in this study for their selflessnessand dedicating time without which the study could not have been a success. Thanks to the personnel atCORDIO, Fisheries Department, and COMRED among other institutions for facilitation.

Funding

The research was funded by Linnaeus University through COMARIO grant and WIOMSA throughgrant MASMA CR/2008/02.

ORCID

Innocent Ngao Wanyonyi http://orcid.org/0000-0003-0317-5271

References

Alvard, M., McGaffey, E., & Carlson, D. (2015). A method for measuring fishing effort by small-scale Fish Aggregating Device (FAD) Fishers from the commonwealth of Dominica [-2015/08//].Field Methods, 27(3), 300–315. doi:10.1177/1525822X14552221

Aswani, S., & Lauer, M. (2006). Incorporating fishermen’s local knowledge and behavior intogeographical information systems (GIS) for designing marine protected areas in Oceania.Human Organization, 65(1), 81–102. doi:10.17730/humo.65.1.4y2q0vhe4l30n0uj

Bastardie, F., Nielsen, J. R., Ulrich, C., Egekvist, J., & Degel, H. (2010). Detailed mapping of fishingeffort and landings by coupling fishing logbooks with satellite-recorded vessel geo-location.Fisheries Research, 106(1), 41–53. doi:10.1016/j.fishres.2010.06.016

Binet, T., Failler, P., & Thorpe, A. (2012). Migration of Senegalese fishers: A case for regionalapproach to management. Maritime Studies, 11(1), 1–14. doi:10.1186/2212-9790-11-1

Brown, G., Schebella, M. F., &Weber, D. (2014). Using participatory GIS to measure physical activity andurban park benefits. Landscape and Urban Planning, 121, 34–44. doi:10.1016/j.landurbplan.2013.09.006

Campbell, M. S., Stehfest, K. M., Votier, S. C., & Hall-Spencer, J. M. (2014). Mapping fisheries formarine spatial planning: Gear-specific vessel monitoring system (VMS), marine conservation andoffshore renewable energy. Marine Policy, 45, 293–300. doi:10.1016/j.marpol.2013.09.015

Cassels, S., Curran, S. R., & Kramer, R. (2005). Do migrants degrade coastal environments?Migration, natural resource extraction and poverty in North Sulawesi, Indonesia. HumanEcology, 33(3), 329–363. doi:10.1007/s10745-005-4142-9

16 I. N. WANYONYI ET AL.

Cinner, J. E., McClanahan, T. R., & Wamukota, A. (2010). Differences in livelihoods, socioeconomiccharacteristics, and knowledge about the sea between fishers and non-fishers living near and far frommarine parks on the Kenyan coast. Marine Policy, 34(1), 22–28. doi:10.1016/j.marpol.2009.04.003

Curran, S. (2002). Migration, social capital, and the environment: Considering migrant selectivityand networks in relation to coastal ecosystems. Population and Development Review, 28, 89–125.

Dunn, C. E. (2007). Participatory GIS — A people’s GIS? Progress in Human Geography, 31(5), 616–637. doi:10.1177/0309132507081493

Forcada, A., Valle, C., Sánchez-Lizaso, J. L., Bayle-Sempere, J. T., & Corsi, F. (2010). Structure andspatio-temporal dynamics of artisanal fisheries around a Mediterranean marine protected area.ICES Journal of Marine Science: Journal Du Conseil, 67(2), 191–203. doi:10.1093/icesjms/fsp234

Fulanda, B., Munga, C., Ohtomi, J., Osore, M., Mugo, R., & Hossain, M. Y. (2009). The structureand evolution of the coastal migrant fishery of Kenya. Ocean & Coastal Management, 52(9), 459–466. doi:10.1016/j.ocecoaman.2009.07.001

Furletti, B., Cintia, P., Renso, C., & Spinsanti, L. (2013). Inferring human activities from GPS tracks.Paper presented at the 2nd ACM SIGKDD International Workshop on Urban Computing.

Gezelius, S. S. (2007). The social aspects of fishing effort. Human Ecology, 35(5), 587–599.doi:10.1007/s10745-006-9096-z

Glaesel, H. (1997). Fishers, Parks and Power: The Socio-environmental dimensions of marine resourcedecline and protection on the Kenya coast. (PhD), University of Wisconsin, Madison, WI.

Gonzalez, M. C., Hidalgo, C. A., & Barabasi, A.-L. (2008). Understanding individual humanmobility patterns. Nature, 453(7196), 779–782. doi:10.1038/nature06958

Government of Kenya. (2007). The Fisheries (Beach Management Unit) Regulations (under TheFisheries Act Cap 378) (Revised 2007).

Government of Kenya. (2012). The Fisheries Act Cap 378 1999 (Revised 2012).Islam, M. M., & Herbeck, J. (2013). Migration and translocal livelihoods of coastal small-scale fishers

in Bangladesh. Journal of Development Studies, 49(6), 832–845. doi:10.1080/00220388.2013.766719Kimani, P., & Obura, D. O. (2007). Participatory mapping of terrestrial fishery resources in Kwale District,

Kenya.Western Indian Ocean Journal of Marine Science, 3(2), 209–220. doi:10.4314/wiojms.v3i2.28462Léopold, M., Guillemot, N., Rocklin, D., & Chen, C. (2014). A framework for mapping small-scale

coastal fisheries using fishers’ knowledge. ICES Journal of Marine Science: Journal Du Conseil, 71(7), 1781–1792. doi:10.1093/icesjms/fst204

Lokrantz, J., Nyström, M., Norström, A. V., Folke, C., & Cinner, J. E. (2009). Impacts of artisanalfishing on key functional groups and the potential vulnerability of coral reefs. EnvironmentalConservation, 36(04), 327–337. doi:10.1017/S0376892910000147

Malleret-King, D., Glass, A., Wanyonyi, I., & Pomeroy, B. (2006). Socio-economic monitoring guidelinesfor coastal managers of the Western Indian Ocean (Vol. 1). Mombasa, Kenya: CORDIO East Africa.

Malleret-King, D., Manghubhai, S., Tunje, J., Muturi, J., Mueni, E., & On’ganda, H. (2003).Understanding fisheries associated livelihoods and the constraints to their development inKenya and Tanzania – Annex 4: Livelihood Appraisal (FANRM/MKK/MRAG), London, UK.

Mangi, S. C., Smith, S., & Catchpole, T. L. (2016). Assessing the capability and willingness ofskippers towards fishing industry-led data collection. Ocean & Coastal Management, 134(SupplementC), 11–19. doi:10.1016/j.ocecoaman.2016.09.027

McClanahan, T. R., Glaesel, H., Rubens, J., & Kiambo, R. (1997). The effects of traditional fisheriesmanagement on fisheries yields and the coral-reef ecosystems of southern Kenya. EnvironmentalConservation, 24(2), 105–120. doi:10.1017/S0376892997000179

McClanahan, T. R., & Mangi, S. C. (2004). Gear-based management of a tropical artisanal fisherybased on species selectivity and capture size. Fisheries Management and Ecology, 11(1), 51–60.doi:10.1111/j.1365-2400.2004.00358.x

McClanahan, T. R., Muthiga, N., & Abunge, C. (2016). Establishment of community managedfisheries’ closures in Kenya: Early evolution of the Tengefu movement. Coastal Management, 44(1), 1–20. doi:10.1080/08920753.2016.1116667

Nielsen, J. R., Degnbol, P., Viswanathan, K. K., Ahmed, M., Hara, M., & Abdullah, N. M. R. (2004).Fisheries co-managementan institutional innovation? Lessons from South East Asia and SouthernAfrica. Marine Policy, 28(2), 151–160.

HUMAN DIMENSIONS OF WILDLIFE 17

Obura, D. O., & Wanyonyi, I. (2001). The local geography of an artisanal fishery and its relevance tofisheries management. Paper presented at the regional fisheries information system project work-shop on the use of information in Tanzanian co-management projects, Tanga.

Ochiewo, J. (2004). Changing fisheries practices and their socioeconomic implications in SouthCoast Kenya. Ocean & Coastal Management, 47, 389–408. doi:10.1016/j.ocecoaman.2004.07.006

Pet-Soede, C., Van Densen, W. L. T., Hiddink, J. G., Kuyl, S., & Machiels, M. A. M. (2001). Canfishermen allocate their fishing effort in space and time on the basis of their catch rates? Anexample from Spermonde Archipelago, SW Sulawesi, Indonesia. Fisheries Management andEcology, 8(1), 15–36. doi:10.1046/j.1365-2400.2001.00215.x

Proud, R., Browning, P., & Kocak, D. M. (2016, 19-23 September). AIS-based mobile satellite serviceexpands opportunities for affordable global ocean observing and monitoring. Paper presented at theOCEANS 2016 MTS/IEEE Monterey.

Rodgers, A. R. (2001). Tracking animals with GPS. Macaulay Land Use Research Institute,Aberdeen. Retrieved from http://www.invasive.org/gist/networks/western/resources/GPS1.pdf

Ron, W. J., Muhando, C. A., & Francis, J. (1998). The status of the coral reefs of Zanzibar: Oneexample of a regional predicament. Ambio, 27(8), 700–707. doi:10.2307/4314818

Rosendo, S., Jiddawi, N., Joubert, A., Mechisso, M., & Brown, K. (2008). Stakeholder needs,perceptions and priorities in relation to use and conservation of natural resources. Lisbon,Portugal: TRANSMAP.

Samoilys, M. A., Osuka, K., Maina, G. W., & Obura, D. O. (2017). Artisanal fisheries on Kenya’scoral reefs: Decadal trends reveal management needs. Fisheries Research, 186(Part 1), 177–191.doi:10.1016/j.fishres.2016.07.025

Scholz, A. J., Steinback, C., Kruse, S. A., Mertens, M., & Silverman, H. (2011). Incorporation ofspatial and economic analyses of human-use data in the design of marine protected areas.Conservation Biology, 25(3), 485–492. doi:10.1111/j.1523-1739.2010.01626.x

Silvano, R. A., & Begossi, A. (2012). Fishermen’s local ecological knowledge on SoutheasternBrazilian coastal fishes: Contributions to research, conservation, and management. NeotropicalIchthyology, 10(1), 133–147. doi:10.1590/S1679-62252012000100013

Sipe, N., & Dale, P. (2003). Challenges in using geographic information systems (GIS) to under-stand and control malaria in Indonesia. Malaria Journal, 2. doi:10.1186/1475-2875-2-36

St. Martin, K. (2001). Making space for community resource management in fisheries. Annals of theAssociation of American Geographers, 91(1), 122–142. doi:10.1111/0004-5608.00236

St. Martin, K. (2006). Charting fishing communities at sea: Revealing new potentials for participation infisheries science and management. Paper presented at the ICES 2006, Fishing Technology in the 21stCentury, Boston, MA. http://scholar.googleusercontent.com/scholar?q=cache:c4kowf9UDA0J:scholar.google.com/+Charting+Fishing+Communities+at+Sea:+Revealing+New+Potentials+for+Participation+in+Fisheries+Science+and+Management+Kevin+St.+Martin&hl=en≈sdt=0,5

St. Martin, K., & Hall-Arber, M. (2008). The missing Layer: Geo-technologies, communities, and implica-tions for marine spatial planning.Marine Policy, 32(5), 779–786. doi:10.1016/j.marpol.2008.03.015

Stone, A. A., & Shiffman, S. (2002). Capturing momentary, self-report data: A proposal for reportingguidelines. Annals of Behavioral Medicine, 24(3), 236–243. doi:10.1207/s15324796abm2403_09

TopoGrafix. (2016). Easy GPS.Van der Spek, S., Van Schaick, J., De Bois, P., & De Haan, R. (2009). Sensing human activity: GPS

tracking. Sensors, 9(4), 3033–3055. doi:10.3390/s90403033Wall, J., Wittemyer, G., Klinkenberg, B., LeMay, V., & Douglas-Hamilton, I. (2013). Characterizing

properties and drivers of long distance movements by elephants (Loxodonta africana) in theGourma, Mali. Biological Conservation, 157, 60–68. doi:10.1016/j.biocon.2012.07.019

Wanyonyi, I. N., Wamukota, A., Mesaki, S., Guissamulo, A. T., & Ochiewo, J. (2016). Artisanalfisher migration patterns in coastal East Africa. Ocean and Coastal Management, 119, 93–108.doi:10.1016/j.ocecoaman.2015.09.006

Wanyonyi, I. N.,Wamukota, A., Tuda, P.,Mwakha, V.A., &Nguti, L.M. (2016).Migrantfishers of Pemba:Drivers, impacts and mediating factors.Marine Policy, 71, 242–255. doi:10.1016/j.marpol.2016.06.009

Yates, K. L.,& Schoeman,D. S. (2013). Spatial Access PriorityMapping (SAPM)withfishers:AquantitativeGIS method for participatory planning. PLoS ONE, 8(7), 1–14. doi:10.1371/journal.pone.0068424

18 I. N. WANYONYI ET AL.