Quantification of carbon stock and tree

diversity of homegardens in a dry zone area of

Moneragala district, Sri Lanka

Eskil Mattsson, Madelene Ostwald, S.P. Nissanka and D.K.N.G. Pushpakumara

Linköping University Post Print

N.B.: When citing this work, cite the original article.

The original publication is available at www.springerlink.com:

Eskil Mattsson, Madelene Ostwald, S.P. Nissanka and D.K.N.G. Pushpakumara, Quantification

of carbon stock and tree diversity of homegardens in a dry zone area of Moneragala district, Sri

Lanka, 2015, Agroforestry Systems, (89), 3, 435-445.

http://dx.doi.org/10.1007/s10457-014-9780-8

Copyright: Springer Verlag (Germany)

http://www.springerlink.com/?MUD=MP

Postprint available at: Linköping University Electronic Press

http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-113524

1

Quantification of carbon stock and tree diversity of homegardens

in a dry zone area of Moneragala District, Sri Lanka†

Eskil Mattssona*, Madelene Ostwalda,b, S.P. Nissankac , D.K.N.G. Pushpakumarac,

a Division of Physical Resource Theory, Department of Energy and Environment, Chalmers University

of Technology, 412 96 Göteborg, Sweden. E-mail: eskil.mattsson@chalmers.se Tel: +46 (0)31 772

2147.

b Centre for Climate Science and Policy Research, Department of Water and Environmental Studies,

Linköping University, 601 74 Norrköping, Sweden.

c Department of Crop Science, Faculty of Agriculture, University of Peradeniya, Peradeniya, Sri

Lanka.

*Corresponding author

†Published online in Agroforestry Systems (Springer Verlag) 31 Dec 2014. The final publication is

available at Springer via http://link.springer.com/article/10.1007/s10457-014-9780-8

Abstract Homegarden agroforestry systems are suggested to hold a large potential for climate change mitigation

and adaptation. This is due to their multifunctional role in providing income, food and ecosystem

services while decreasing pressure on natural forests and hence saving and storing carbon. In this

paper, above-ground biomass carbon and tree species diversity of trees was quantified in homegardens

around two villages in the dry south-eastern part of Moneragala district of Sri Lanka. A total of 45 dry

zone homegardens were sampled on size, diameter at breast height, tree height and species diversity.

Using allometric equations, we find a mean above-ground biomass stock of 13 Mega grams of carbon

per hectare (Mg C ha-1) with a large range among homegardens (1 to 56 Mg C ha-1, n=45) due to a

variation of tree diversity and composition between individual homegardens. Mean above-ground

carbon stock per unit area was higher in small homegardens (0.2 ha, 26 Mg C ha-1, n=11) and

statistically different compared to medium (0.4–0.8 ha, 9 Mg C ha-1, n=27) and large (1.0–1.2 ha, 8

Mg C ha-1, n=7) homegardens. In total, 4,278 trees were sampled and 70 tree species identified and

recorded. The Shannon Wiener index were used to evaluate diversity per homegarden and ranged from

0.76 to 3.01 with a mean value of 2.05 0.07 indicating a medium evenly distributed diversity of

sampled tree species. The results show a vast heterogeneity in terms of carbon stock and tree diversity

within the less studied dry zone homegardens; results that contribute to more knowledge of their

expansion potential as well as climate mitigation and adaptation potential. The results are also useful

for whether homegardens should be considered to be included as an activity to enhance natural forest

cover within Sri Lanka’s newly commenced UN-REDD National Programme.

Keywords: allometric equations; biomass; carbon; agroforestry, REDD+

2

Introduction Homegardens are agroforestry systems common throughout the tropics (Nair and Kumar 2006; Mohri

et al. 2013). Homegardens are prime examples of multi-functional landscapes: spaces that combine

agriculture, forestry and natural ecosystems and are in Sri Lanka defined as a complex sustainable

land-use system that combines multiple farming components, such as annual and perennial crops,

livestock and occasionally fish, of the homestead and provides environmental services, household

needs, and employment and income generation opportunities to the households (Weerahewa et al.

2012). Others, such as, Kumar and Nair (2004) and Nair and Kumar (2006) provide a similar

definition and suggest that a homegarden includes the concept of intimate plant associations of various

trees and crops, sometimes in association with domestic animals, and consequent multi-story canopy

configuration around the homestead. Despite the lack of a uniform definition of homegardens they are

second to shifting cultivation the oldest land-use systems worldwide that have evolved through

centuries of biological and cultural transformation (Nair and Kumar 2006; Pushpakumara et al. 2012).

In recent years, there has been growing interest in agroforestry systems due to their large potential for

climate change mitigation and adaptation and their role to mitigate household food security and

nutrition from soaring food prices (Minang et al. 2012; Nair 2012; Galhena et al. 2013).

Homegardens also store higher amounts of carbon than other agriculture systems in the above- and

below-ground biomass and soils, but usually inferior to mature forests at the same site (Schroth et al.

2011; Mattsson et al. 2013). The provisioning role of agroforestry and homegardens to maintain

species diversity may also facilitate more stable and longer term stability of carbon stocks as well as

diversification of homegarden derived products (Yachi and Loreau 1999; Brookfield et al. 2002;

Henry et al. 2009).

To meet future challenges of land and water scarcity, and to ensure food security as a result to adverse

effects of climate change, future mitigation and adaptation strategies that can be used by local land

users through effective support by stakeholders and policymakers needs further attention (Murthy et

al. 2013). To identify such strategies, it is relevant to analyze quantitative information and estimates of

tropical homegardens’ ability to sequester and store carbon. Although, the importance and recognition

of homegardens for carbon storage has been highlighted earlier (e.g., Kumar 2006; Nair 2012) there is

still a lack of quantitative data on homegardens and their carbon content, especially in dry zone

environments in Sri Lanka. Few studies have also related species diversity to ecosystem processes

(Pushpakumara et al. 2012). Since subsistence agriculture is predominantly practiced in the dry zone,

the little research focus on dry zone homegardens warrants further investigation on this subject.

Therefore, this paper focuses on assessing the amount and pattern of tree diversity and above-ground

biomass (AGB) in homegardens around two selected villages in the southern part of Moneragala

district in Sri Lanka. The objective is to investigate how the AGB carbon and tree diversity varies

depending on parameters such as homegarden size, soil organic carbon (SOC) content, diameter at

breast height (DBH) and tree height.

Carbon stocks and tree diversity in Sri Lankan homegarden systems

Based on the mean annual rainfall, Sri Lanka is divided into three climatic zones. They are dry,

intermediate, and wet mean annual rainfall of less than 1,750, between 1,750 to 2,500 and over 2,500

mm, respectively (see Fig. 1) (MFE 1995). In Sri Lanka, homegardens cover about 14% of the total

land area (FAO 2009). They are mainly privately owned and are managed through family labor using

technologies that rely on rich local knowledge systems (Pushpakumara et al. 2010). Sri Lankan

homegardens provide staple foods, many kinds of fruits, vegetables, spices, fuelwood, fodder, timber,

medicinal plants and occasionally livestock products throughout the year. Although being ancient

3

land-use systems, governmental policies have recently encouraged establishment and expansion of

homegardens in urban and sub-urban settings as a mechanism to reduce living costs and imports of

food products, to enhance food security and maintenance of environmentally friendly-traditional

agriculture methods. Examples include the most recent program “Divi Neguma” (Livelihood

Development) that aim to establish 2.5 million homegardens to achieve self-sufficiency in vegetables

and to reduce vegetable prices by providing free seeds, fertilizer and technical advice (Kumari et al.

2009; Government of Sri Lanka 2011). So far homegarden research in Sri Lanka has predominantly

included various assessments of Kandyan homegardens located in the mid-country wet zone

(Pushpakumara et al. 2010).

Few comparative analyses exists both in terms of tree diversity and AGB carbon stocks for dry zone

environments in Sri Lanka. However, Mattsson et al (2013) estimated the carbon stock for dry zone

Sri Lankan homegardens (Hambantota and Anuradnapura district) is ranging from 10 to 55 Mega

grams of carbon per hectare (Mg C ha-1) (mean 35 Mg C ha-1). For comparison, Kandyan homegardens

in the wet zone had carbon stocks ranging between 48 to 145 Mg C ha-1 with a mean value of 87 Mg C

ha-1 (ibid.)

Homegarden diversity is by large controlled by ecological and socio-economic factors such as altitude,

homegarden size, age of gardens, and personal preferences by the gardeners, market access and

production intensity (Karyono 1990; Abdoellah et al. 2006; Nair and Kumar 2006; Peyre et al. 2006;

Torquebiau and Penot 2006; Wiersum 2006; Pandey et al. 2007; Kehlenbeck and Maas 2008).

Ariyadasa (2002) estimated more than 400 different woody species in Sri Lankan homegardens. In 20

districts a total of 153,493 million trees have been recorded of which wet, intermediate and dry zone

homegardens consists of 49, 37, and 14% of trees, respectively. The average density of trees in

homegardens of Sri Lanka is varying from 20 to 475 trees ha-1. The wet zone homegardens are

considerably smaller in extent than dry zone homegardens but the total number of trees recorded

within the wet zone is much higher due to the higher tree density (average of 260 trees ha-1) than those

in the dry zone (average 125 trees ha-1). Higher species density has also been found in small

homegardens than large homegardens (McConnell and Dharmapala 1973; McConnell 2003;

Pushpakumara et al. 2012).

Material and methods

Study site

The study was conducted in the neighboring villages of Padikapuhela (6°21’30’ N–6°22’30’N and

81°14’30’ E–81°15’10’ E) and Pilimihela (6°21’00’ N–6°21’60’ N and 81°15’20’ E–81°15’60’ E),

located in Thanamalwila Division in the most southern part of Moneragala district in the south-eastern

part of Sri Lanka (Fig. 1). The topography is undulating lying between 50–70 meters above sea level.

Together with three small villages of same size, Padikapuhela and Pilimihela belong administratively

to Beralihela which had a population of 1,793 in 2001. The villages are located within the driest agro-

ecological zone (DL1) in the country where the soils are predominately alfisols (Panagos et al. 2011).

The mean annual temperature is 27 °C and the mean annual rainfall is 1,050 mm (Department of

Census and Statistics 2010). Most rain falls during the north-east monsoon (Maha season) from

October to January while the south-west monsoon (Yala season) lasts from May to August and bring

less rain (Seo et al. 2005). Subsistence agriculture from paddy rice, banana, homegardens and shifting

cultivation is the main income sources for the villagers (Withanage, pers. comm.). The village was

established during the Kirindi Oya Irrigation and Settlement Project (KOISP) in 1979–1986 (Nijman

4

1992) where the government established new irrigated land within the Kirindi Oya River drainage

basin to increase of food production, employment and providing arable land (IIMI 1995). Hence,

natural water flows are scarce in the area but diversion of irrigated water from Lunugamvehera

reservoir provides water sufficiently for crop cultivation of paddy and banana during both monsoonal

seasons. The area is also located close to the Lunugamvehera National Park, and thus the area is prone

to wildlife interaction with negative effects on yields. Recently, intensive infrastructure development

including an international airport and road developments has taken place close to the study area

through state and foreign aided funds with increased prices of property as a result. A tank renovation

project has also lately been established in the village through a Global Environmental Facility (GEF)

Small Grants community based adaptation project and is expected to increase the water table and

provide enough water for agricultural practices in homegardens and shifting cultivation lands for the

villagers in the dry seasons.

Fig.1 The study area and its location in the dry zone of the most southern part of Moneragala district

as highlighted in red on the left image and green as sampling locations of homegardens on the right

image. Note: 1, 2 and 3 of the left image represent the dry, intermediate and wet zones. Source:

Modified from Google Earth, V 7.1.1.1888. 24 February (2013).

Mapping and data analysis of homegardens

The selected villages of Padikapuhela and Pilimihela were chosen, partly because the villages were

included in a larger study in a Global Environmental Facility (GEF) Small Grants community based

adaptation project “Developing community-led strategies and infrastructure to ensure adaptation to

drought conditions”. The objective of the programme was to ensure sustainable agricultural practices

for rainfed farming families through village tank development, training programmes, marketing

mechanism for agricultural products and introduction of alternate livelihoods (AUSAid 2012).

5

Forty-five homegardens were randomly sampled in March 2013. To cover the AGB carbon variation

and species diversity of the individual homegardens in the village, the total area of all homegardens

were sampled which resulted in a total sample area of 30 hectares. The homegardens were randomly

selected in order to capture a representative mixture of size of homegardens, diversity and composition

of tree species, as well as socio-economic status of householders although an analysis of the latter is

not assessed in this paper. All sampled homegardens were between 20–30 years old and the size of

homegardens ranged from 0.2–1.2 ha where a majority of homegardens were rectangular in shape.

Data on SOC collected from the same homegardens in March 2013 (n=40; Granberg et al. submitted)

was used for data analysis to explore the correlation of SOC and AGB carbon. Data collected were

subjected to analysis with SPSS version 21 using descriptive statistics, correlation analysis and

stepwise multiple regression analysis (OLS).

Above-ground biomass carbon

All perennial trees and plants with a diameter at breast height (DBH) of ≥ 3 cm were measured. The

DBH and tree height were measured using a DBH tape and a clinometer, respectively. Species

information was collected for each tree in each homegarden. For comparison reasons, the

homegardens were categorized into three size groups, namely small (0.2 ha), medium (0.4–0.8 ha) and

large homegardens (1.0–1.2 ha). Given the lack of a standard approach and available allometric

equations to estimate AGB for homegarden agroforestry systems in Sri Lanka, pan-tropical allometric

equations developed for tropical natural forests were used. The allometric equation developed by

Chave et al (2005) for dry zone forests (rainfall under 1,500 mm yr-1) was applied for individual trees

using measured DBH and tree height as well as literature derived species-specific wood density of all

sampled species (varying from 0.26–1.06 g cm3) following Reyes et al (1992) and the Wood Density

Database by Chave et al (2009) as input variables. In those cases wood densities were not found, a

default value of 0.57 g cm-3 for the Asian region was used (Reyes et al. 1992). For bananas and palms

(primarily for Cocos nucifera) we used the equations developed by Arafin (2001) (cited in Hairiah et

al. 2010) and Brown (1997), respectively (Table 1).

Table 1 Allometric equations used to estimate above-ground biomass for individual trees, bananas and

palms in the dry zone homegardens

Type of above-ground

biomass

Allometric equation R2 Source

Individual trees

Y = exp (-2.187+ 0.916 ×

ln (D2×H×S))

0.99

Chave et al. 2005

Bananas Y = 0.030 D2.13 0.99 Hairiah et al. 2010

Palms Y = exp (-2.134 + 2.530 x

ln(D)}

0.97 Brown 1997

Y = above-ground biomass density (Mg ha-1), D = diameter in cm, H = height in m, S = species-specific wood

density in g cm-3.

Tree biomass was converted to carbon assuming that carbon accounted for 50% of the biomass

(Brown 1997). The AGB carbon stock was calculated for each tree and was aggregated to calculate

total AGB carbon stock for each homegarden. For comparisons on unit area basis the values were

extrapolated to hectare size.

6

Tree diversity

Tree species diversity was assessed within the fixed boundaries of the sampled homegardens acquiring

common names that subsequently was translated into botanical names. An index was set up based on

the number of species and their frequency in the homegardens. For this study we chose to use the

Shannon Wiener index (SWI) due to its suitability for evaluating diversity in carbon sequestration

projects (Ponce-Hernandez 2004). The Shannon-Weiner diversity characterizes the proportion of

species abundance in the population, being at maximum when all species are equally abundant and the

lowest when the sample contains one species. Shannon Diversity Index values usually range between

1.5 and 3.5 and seldom more than 4.5 (ibid). The proportion of species (i) relative to the total number

of species ( ) was calculated and then multiplied by the natural logarithm of the same proportion (Ln

). The resulting product is summed across species, and multiplied by -1 (equation 1).

ln( )

Results

Above-ground biomass carbon

The above-ground biomass (AGB) carbon stock for the 45 sampled homegardens ranged from 1.0 to

56.7 Mg C ha-1 with a mean value of 12.7 Mg C ha-1. Per unit area basis, mean AGB carbon stock was

higher in small homegardens (0.2 ha, 26 Mg C ha-1, n=11) and statistical different (p< 0.05) compared

to medium (0.4–0.8 ha, 9 Mg C ha-1, n=27) and large (1.0–1.2 ha, 8 Mg C ha-1, n=7) homegardens

(Fig. 2). By comparing the AGB carbon stock with number of tree species (Fig. 3) it is shown that that

there is positive trend (not statistically significant) in relationship for small homegardens, a less

positive relationship for medium size homegardens and a small negative relationship for large

homegardens.

7

Fig. 2 Above ground biomass (AGB) carbon stock of small (0.2 ha), medium (0.4–0.8 ha) and large

(1.0–1.2 ha) dry zone homegardens in Moneragala district of Sri Lanka. Error bars show standard

error. Small homegardens are statistically different compared to medium and large sized homegardens

whereas no significant differences are found between medium and large size homegardens.

Fig. 3 The relationship between above ground biomass (AGB) carbon stock per unit area and the

diversity of trees of small (0.2 ha), medium (0.4–0.8 ha) and large (1.0–1.2 ha) dry zone homegardens

in Moneragala district of Sri Lanka.

Tree diversity

In total, 70 different tree and plant species were identified from 55 genera and 30 families and in total

4,278 trees were measured (see Online Resource 1). Fourteen species were unidentified. The most

common species found was Neem (Azadirechta indica A. Juss., n=1014) accounting for 24% of all

trees measured followed by Cashew (Anacardium occidentale L., n=509, 12%) and Coconut (Cocos

nucifera L., n=362, 8%). Tree diversity described by the SWI showed a variation between 0.76–3.01

with a mean value of 2.05 0.07 where small sized homegardens had the highest mean diversity of

trees, followed by medium and large homegardens (Table 2). Mean number of tree species ha-1 were

80, 23 and 15, respectively for small, medium and large sized homegardens.

8

Table 2 Tree diversity characteristics in different categories of homegardens of southern dry zone

areas of Moneragala district in Sri Lanka.

Homegarden size

Species

recorded per

homegarden

Shannon Wiener index

(SWI)

Mean number

of species ha-1

Total Mean Mean SE Range

Small (0.2 ha)

n=11

43 14.91 2.08 0.15 0.76 –

2.66 80 10.03

Medium (0.2–0.8 ha)

n=27

58 15.52 2.07 0.09 1.08 –

3.01 23 2.13

Large (1.0–1.2 ha)

n=7

37 15.57 1.94 0.16 1.52 –

2.60 15 1.78

All categories

73 15.38 2.05 0.07 0.76 –

3.01 36 4.65

SE standard error

Species area curve for the selected villages showed that covering of 90% of tree species require

roughly 15 ha of cumulative land area or 50% of sampled area. The maximum diversity of an

individual garden was recorded in a medium size garden and the minimum diversity was found in a

small homegarden (Table 2). As the size of homegardens increased, species richness on total

homegarden area basis showed a very small increase. However, species richness ha-1 was much greater

in small sized homegardens and secondly, medium sized homegardens than the large homegardens

(Fig. 4).

No statistical significant difference was found between size-groups in terms of total species using two-

tailed exact Mann-Whitney U test even though the trend showed that the small and medium sized

groups were more homogenous while the large size group showed a greater difference.

9

Figure 4 Tree species richness and tree species richness per hectare (ha) in relation to homegarden

size for each sampled homegardens. Each homegarden is represented by a blue and a red symbol

(2x40 homegardens = 80 symbols). Linear and exponential regression lines with equations and r-

squared values are shown for species richness and for species richness per ha, respectively.

Correlation between carbon and tree diversity

A correlation analysis (a non-parametric Spearman test) was conducted by using soil organic carbon

(SOC) measures from 40 of the same homegardens in a companion study (Granberg et al. submitted)

where carbon content at 0–15 cm and 15–30 cm depth were assessed. No significant correlation

between the SOC components and the terrestrial carbon components was found in our assessment (i.e.,

AGB, trees ha-1, species ha-1, size of homegardens, DBH or tree height). Hence, the SOC data was

removed for further analyses. The remaining terrestrial carbon component did show significant

correlation in several cases (Table 3). Since some of the parameters are directly correlated, such as

size and trees or species ha-1, it is worth noting that AGB correlate positively and significantly with

trees and species ha-1 and height, while DBH shows no correlation. Further, the significant negative

correlation between AGB and homegarden size suggests that the larger the homegarden the less

carbon does the homegarden contain per unit area.

10

Table 3 Non-parametric correlation (Spearman’s Rho) n=45

Spearman’s Rho AGB Trees ha-1 Species ha-1 Size DBH

Trees ha-1 0.748**

Species ha-1 0.718** 0.814**

Size –0.57** –0.601** –0.741**

DBH 0.224 –0.302* –0.209 0.253

Height 0.436** 0.271 0.346 –0.221 0.211

** correlation is significant at the 0.01 level (2-tailed);

* correlation is significant at the 0.05 level (2-tailed)

Discussion Firstly, the large variability of AGB carbon among homegardens are primarily a result of tree density

which is highest in small homegardens (small: 80 trees ha-1), followed by medium (23 trees ha-1) and

larger units (15 trees ha-1). This is consistent with earlier findings (e.g. Kumar 2011, Pushpakumara et

al. 2012) although tree density is lower than those dry zone homegardens assessed in the study by

Ariyadasa (2002). Secondly, differences in carbon stock is a result of differences in tree diversity,

management practices, homegarden age, site characteristics and composition differences (Montagnini

and Nair 2004; Henry et al. 2009). For example, in the large homegardens investigated in this study,

farmers often use shifting cultivation practices to prepare and burn certain plots within the gardens for

cash crop cultivation in the wet season with less tree densities as a result. Excessive shading from trees

is also thought to have a detrimental effect on cash crop yields according to the homegarden owners.

As stressed earlier, few comparative analyses exists both in terms of tree diversity and AGB carbon

stocks for homegardens in dry zone environments in Sri Lanka. However, the mean AGB carbon (13

Mg C ha-1) is lower and the large range (1 to 56 Mg C ha-1) between individual homegardens is higher

than values shown earlier from Mattsson et al (2013). Dissanayake et al (2009) estimated the AGB

carbon stock in homegardens in Kandy (90 Mg C ha-1) and Matale (104 Mg C ha-1) districts and

Premakantha et al (submitted) reported that homegardens in Nuwara Eliya district contain 77 Mg C ha-

1. All the latter estimates are located within the wet zone and intermediate zone, and as expected, these

values are higher due to a higher level of plant diversity and a denser canopy structure than dry zone

homegardens. Outside Sri Lanka, Kumar (2011) found that Kerala homegardens in India had AGB

carbon stocks ranging from 16 to 36 Mg ha−1. Estimates from Javanese and Sumatran homegardens

(35 to 59 Mg C ha−1; Jensen 1993; Roshetko et al. 2002) as well as mature (>35-year old) agroforests

(101 Mg C ha−1) are higher than our estimates, but similar to 9-year old Sumatran agroforests (14 Mg

C ha−1; Roshetko et al. 2002).

For tree diversity, the mean Shannon Wiener Index (SWI) of 2.05 is lower than the SWI found by

APN (2012) in the homegardens of the Keeriyagaswewa village (SWI: 2.13; n=59) located in the Sri

Lankan dry zone but higher than in Siwalakulama village (SWI: 1.77; n=30; intermediate zone) and

Pethiyagoda village (SWI: 1.99; n=59, wet zone) (ibid.) and in the Meegahakiula area (SWI: 1.55 to

1.77; intermediate zone; Senanayake et al. 2009). The estimated SWI is also higher than the mean

SWI found in homegardens from two villages in West Bengal, India and six villages in Dhaka

Division, Bangladesh (APN, 2012), but lower than in Kerala homegardens in India (Saha et al. 2009).

Mean Shannon-Wiener diversity indices in tropical homegardens have been reported to vary broadly

from 0.93 in rural Zambia (Drescher 1998) to almost 3.0 in West Java, Indonesia (Karyono 1990).

11

Thus, the Shannon index is in keeping with prior studies of homegardens studied in other regions of

Sri Lanka, indicating a medium evenly distributed diversity. This study also only considers trees, and

the overall SWI would have been higher if all plants were included. The negative relationship between

tree species richness and large homegardens is in keeping with Kumar (2011) although e.g. Abdoellah

et al (2006) and Mendez et al (2001) found low correlation whereas Abebe (2005), Kabir and Webb

(2009) and Bardhan et al (2012), found a positive relationship. Our negative relationship between tree

species richness and large homegardens is probably a result of specific needs and preferences by the

owner along with socio-economic factors. For example, in larger homegardens there are often distinct

spatial arrangements between trees and crops, where most trees are planted around the house and low

tree density areas are found towards the borders where cash crops such as bananas or leguminous

crops are grown for imminent needs and local markets, hence lower tree diversity and tree density. In

small homegardens on the other hand, homegarden owners often adopt more intensive management

and denser planting in multiple layers, thus, higher tree species richness.

No relationship was found between measured parameters and soil organic carbon (SOC) in

homegardens measured concurrently (Granberg et al. submitted) in a companion study. This is not

consistent with the findings of Saha et al (2009) who found that species richness could provide greater

stability of the SOC. The positive relationship between the SWI index and AGB carbon stock for

small-sized gardens in relation to the slightly negative relationship for large homegardens and AGB

carbon suggests that small-sized homegardens hold higher carbon stock and species diversity, in line

with Kumar (2011) and Pushpakumara et al (2012). Therefore, expanding homegardens into degraded

and low-productive lands (such as shifting cultivation lands) for increased ecosystem services and

carbon stocks as suggested by Mattsson et al (2013) would thus be more beneficial through the

provision of small homegarden units rather than larger units. Also, the traditional use of homegardens

has been around homesteads inherently meaning close to where people live. For sustained and

enhanced ecosystem services such as carbon storage through expansion, would require new

management strategies, for example that homegarden systems integrate more commercial oriented

example plantation of perennials through a combination of exotic and/or indigenous species.

For the study area per se, recent large-scale infrastructure developments and investments in the area

could in coming years have an effect on land prices, leading to a higher population density and

increased demand for arable land. The village tank development under the auspices of the GEF Small

Grants community based adaptation project are upon finish expected to provide water for both

cultivation seasons (Maha and Yala) in the villages. Along with an increasing population, more

homegardens are likely to be established while water accessibility would allow farmers and

homegarden owners to opt for more perennial plants of different varieties and species which overall

can enhance the carbon density and tree diversity which could favor higher productivity and ultimately

sustained or improved food security.

The results of this study show that the investigated homegardens have good capacity for carbon

storage and sequestration capacity which provides useful information for the national process of

whether homegardens should be considered to be included as an activity within Sri Lanka’s

commenced National Programme on REDD+. This implies that developed countries provide

incentives and financial compensation to developing countries for climate change mitigation benefits

from maintaining and enhancing forest biomass. Homegardens could in this context be considered to

be included within an existing or new forest definition to lower greenhouse gas emissions directly

(Minang et al. 2014). Another option could be to further promote homegarden establishment by

intensification or extensification on marginal lands and serve as important buffers for the remaining

12

natural forests in areas that are experiencing pressure from increasing populations. Consequences at

the household level for such programmes needs to be explored further to investigate possible trade-

offs between for example biodiversity, economic, nutritional and financial benefits. For best outcomes

at the local level, schemes such as REDD+ could be linked to existing or emerging developing

programs highlighting food security and market integration since associated finance coming out of

emissions based implementation is unlikely to be of main concern for local farmers (Bernard et al.

2013).

Conclusions Our results suggests that the investigated homegardens in the southern part of the dry zone of

Moneragala district of Sri Lanka hold a wide range of carbon between 1 to 56 Mg ha-1 and a mean

above-ground biomass stock of 13 Mg carbon ha-1, which is lower than other reported carbon

estimates for homegardens in different ecological zones. The carbon estimates found here are

reflecting the differences in tree density, tree diversity and management practices between individual

homegardens. Furthermore, no correlation is found between soil carbon parameters and AGB carbon

or diversity of trees. Smaller homegardens hold a higher carbon content and tree diversity than

medium and large homegardens. Hence, based on the result here the expansion potential into degraded

lands or larger units is not straight forward if carbon stock and tree diversity should be kept.

Acknowledgements This research was funded by the Swedish Energy Agency and the Gothenburg Centre for Global

Development (GCGD). The authors would like thank Janaka Withanage and Thusitha Madurande

Geeganage for logistical assistance and the farmers in Beralihela who participated in the project.

References

Abebe T. (2005) Diversity in Homegarden Agroforestry Systems of Southern Ethiopia. Wageningen

University, the Netherlands, Tropical Resource Management Paper No. 59.

Abdoellah OS, Hadikusumah HY, Takeuchi K, Okubo S, Parikesit (2006) Commercialization of

homegardens in an Indonesian village: vegetation composition and functional changes. Agroforest

Syst 18:1–13.

APN (Asia-Pacific Network for Global Change Research) (2012) Vulnerability of Home Garden

Systems to Climate Change and its Impacts on Food Security in South Asia. Retrieved from

http://www.apn-gcr.org/resources/items/show/1566

Ariyadasa KP (2002) Assessment of tree resources in the homegardens of Sri Lanka. Bangkok,

ECFAO Partnership Programme on Information and Analysis for Sustainable Forest Management

(FAO, 2005).

AUSAID (2012) AusAID Community Based Adaptation (CBA) Project CPMT Consolidated Report.

39 p.

https://sgp.undp.org/index.php?option=com_sgpprojects&view=projectdetail&id=16559&Itemid=205.

Accessed 2013-12-19.

13

Bardhan S, Jose S, Biswas S, Kabir, K, Rogers, W (2012) Homegarden agroforestry systems: an

intermediary for biodiversity conservation in Bangladesh. Agroforest Syst 85:29–34.

Bernard F, Minang PA, van Noordwijk M, Freeman OE, Duguma LA. (Eds) (2013) Towards a

landscape approach for reducing emissions: A substantive report of Reducing Emissions from All

Land Uses (REALU) project. Nairobi, Kenya. World Agroforestry Centre – ICRAF. 119 p.

Brookfield H, Stocking M, Brookfield M (2002) Guidelines on agrodiversity assessment. In:

Brookfield, H., Padoch, C., Parsons, H., Stocking, M. (Eds.), Cultivating Biodiversity: Understanding,

Analysing and Using Agricultural Diversity. ITDG Publishing, London, UK, pp. 41–56.

Brown S (1997) Estimating biomass and biomass change of tropical forests: a primer. In: Forestry

Paper 134–publication FRA, Rome, FAO.

Chave J, Andalo C, Brown S, Cairns MA, Chambers JQ, Eamus D, Folster H, Fromard, F et al (2005)

Tree allometry and improved estimation of carbon stocks and balance in tropical forests. Oecologia

145:87–89.

Chave J, Coomes DA, Jansen S, Lewis SL, Swenson NG, Zanne AE. (2009) Towards a worldwide

wood economics spectrum. Ecol Lett 12(4): 351–366.

Department of Census and Statistics. (2010) Statistical abstract 2010, Chapter 1 – Area and Climate.

http://www.statistics.gov.lk/abstract2010/Pages/chap1.htm. Accessed 2013-10-23.

Dissanayake WASS, Ranasinghe, DMSHK, Wahala S (2009) Estimation of carbon stock in Kandyan

Homegardens located in Kandy and Matale. In Proceedings of the 14th International Forestry and

Environment Symposium, Department of Forestry and Environmental Science, University of Sri

Jayewardenepura.

Drescher AW (1998) Hausgärten in afrikanischen Räumen: Bewirtschaftung nachhaltiger

Produktionssysteme und Strategien der Ernährungssicherung in Sambia und Zimbabwe. Centaurus-

Verlagsgesellschaft, Pfaffenweiler, Germany, 275 pp.

FAO (2009) Sri Lanka Forestry Outlook Study. Asia Pacific Forestry Outlook Sector Outlook Study

II, Working Paper No. APFSOS II/WP/2009/29, Bangkok, FAO Regional Office for Asia and the

Pacific.

Galhena DH, Freed R, Maredia, KM (2013) Home gardens: a promising approach to enhance

household food security and wellbeing. Agriculture & Food Security, 2013; 2(8):48–62.

Google Earth, V 7.1.1.1888. 24 February (2013) Beralihela, Sri Lanka. 6°21’36.91’’N

81°15’13.42’’E. Eye alt 5.61 km. Digital Globe 2013. http://www.earth.google.com [Nov 1, 2013].

Government of Sri Lanka (2011) Improving family nutrition, a main objective of Divi Neguma

Programme. Retrieved October 11, 2011, from

http://www.priu.gov.lk/news_update/Current_Affairs/ca201107/20110713improving_family_nutrition

.htm.

14

Hairiah K, Dewi S, Agus F, Velarde S, Ekadinata A, Rahayu S, van Noordwijk M (2010) Measuring

Carbon Stocks Across Land Use Systems: A Manual. Bogor, Indonesia. World Agroforestry Centre

(ICRAF), SEA Regional Office, 155 p.

Henry M, Tittonell P, Manlay RJ, Bernoux M, Albrecht A, Vanlauwe B (2009) Biodiversity, carbon

stocks and sequestration potential in aboveground biomassin smallholder farming systems of western

Kenya. Agr Ecosyst Environ 129 (1): 238–252.

IIMI (1995) Kirindi Oya Irrigation Settlement Project. Project Impact Evaluation Study. Colombo:

IIMI (IIMI report 1995:02).

Jensen M (1993) Soil conditions, vegetation structure and biomass of a Javanese homegarden.

Agroforest Syst 24:171–186.

Kabir ME, Webb EL (2009) Household and homegarden characteristics in southwestern Bangladesh.

Agrofor Syst 75:129–145.

Karyono (1990) Home Gardens in Java. Their Structure and Function. In: Landauer K. and Brazil M.

(eds). Tropical Home Gardens. The United Nations University, Tokyo, Japan, pp. 138–146.

Kehlenbeck K, Maass BL (2004) Crop diversity and classification of homegardens in Central

Sulawesi, Indonesia. Agroforest Syst 63:53–62.

Kumar BM (2006) Carbon sequestration potential of tropical homegardens. In Tropical homegardens:

A time-tested example of sustainable agroforestry, ed. BM Kumar, and PKR. Nair, 185–204.

Dordrecht: Springer.

Kumar BM (2011) Species richness and aboveground carbon stocks in the homegardens of central

Kerala, India. Agr Ecosyst Environ 140:430–440.

Kumar BM, Nair PKR (2004) The enigma of tropical homegardens. Agroforest Syst 61:135–152.

Kumari MAS, Kansuntisukmongkol K, Brockelman WY (2009) Plant diversity in home gardens and

its contribution to household economy in suburban areas in Sri Lanka. Environment and Natural

Resources Journal 7:12–30.

Mattsson E, Ostwald M, Nissanka SP, Marambe B (2013) Homegardens as a multi-functional land use

strategy in Sri Lanka with focus on carbon sequestration. AMBIO 42:892–902.

McConnell DJ (2003) The forest farms of Kandy and other gardens of complete design. Ashgate

Studies in Environmental Policy and practice. Ashgate Publishing Limited, England. 370 pp.

McConnell DJ, Dharmapala KAE (1973) The economic structure of Kandyan forest garden farms.

Small forest garden farms in the Kandy district of Sri Lanka. Farm Management Diversification

Report No. 7. UNDP/SF/FAO Diversification Project, Peradeniya, Sri Lanka.

Mendez VE, Lok R, Somarriba E (2001) Interdisciplinary analysis of homegardens in Nicaragua:

micro-zonation, plant use and socioeconomic importance. Agrofor Syst 51:85–96.

15

MFE (1995) Sri Lanka Forestry Sector Master Plan. Forestry Planning Unit, Ministry of Forestry and

Environment, Battaramulla, Sri Lanka, 511 pp.

Minang PA, van Noordwijk M, Swallow BM (2012) High-Carbon-Stock Rural Development

Pathways in Asia and Africa: Improved Land Management for Climate Change Mitigation. In:

Agroforestry: The Future of Global Land Use. Nair PKR and Garrity DP (eds.), Springer, The

Netherlands pp 127–143.

Minang PA, Duguma LA, Bernard FO, Mertz O, van Noordwijk M (2014) Prospects for agroforestry

in REDD+ landscapes in Africa. Curr Opin Environ Sustain 6:78–82.

Mohri H, Lahoti S, Saito O, Mahalingam A, Gunatilleke N, Van Thang Hoang I, Hitinayake G,

Takeuchi K, Herath S (2013) Assessment of ecosystem services in homegarden systems in Indonesia,

Sri Lanka, and Vietnam. Ecosystem Services 5:124–136.

Montagnini F, Nair PKR (2004) Carbon sequestration: An underexploited environmental benefit of

agroforestry systems. Agroforest Syst 61:281–295.

Murthy IK, Gupta M, Tomar S, Munsi M, Tiwari R et al (2013) Carbon Sequestration Potential of

Agroforestry Systems in India. J Earth Sci Climate Change 4(1):1-7. doi:10.4172/2157-7617.1000131

Nair PKR, Kumar BM (2006) Introduction. In Tropical homegardens: A time-tested example of

sustainable agroforestry, ed. B.M. Kumar, and P.K.R. Nair, 1–10. Dordrecht: Springer.

Nair PKR (2012) Climate change mitigation: a low-hanging fruit of agroforestry, in Agroforestry – the

Future of Global Land Use, ed. by Nair PKR and Garrity D. Springer, Dordrecht, pp. 31–67.

Nijman C (1992) Irrigation Decision-Making Processes and Conditions – A Case Study of Sri Lanka´s

Kirindi Oya Irrigation and Settlement Project. Colombo: International Irrigation Management Institute

(IIMI) (IIMI Report, 1992:09).

Panagos P, Jones A, Bosco C, Senthil Kumar PS (2011) European digital archive on soil maps

(EuDASM): Preserving important soil data for public free access. International Journal of Digital

Earth, 4 (5), pp. 434–443.

Pandey CB, Rai RB, Singh L, Singh AK (2007) Home gardens of Andaman and Nicobar, India. Agr

Syst 92(1-3):1–22.

Peyre A, Guidal A, Wiersum KF, Bonger F (2006) Homegarden dynamics in Kerala, India. In:

Tropical Homegardens: A Time-tested Example of Sustainable Agroforestry. Eds. C. Belair, K.

Ichikawa, B.Y.L. Wong and K.J. Mulongoy. pp 102–108. Springer Science, Dordrecht.

Ponce-Hernandez, R (2004) Assessing Carbon Stocks and Modelling Win–win Scenarios of Carbon

Sequestration through Land-use Changes. Food and Agriculture Organization of the United Nations,

Rome, Italy.

16

Pushpakumara DKNG, Wijesekara A, Hunter DG (2010) Kandyan homegardens: a promising land

management system in Sri Lanka. In Sustainable use of biological diversity in socio-ecological

production landscapes, Ed. Belair, C., K. Ichikawa, B.Y.L. and K.J. Mulongoy. Background to the

‘Satoyama Initiative for the benefit of biodiversity and human well-being. Secretariat of the

Convention on Biological Diversity (2010), Montreal. Canada, pp 102-108.

Pushpakumara DKNG, Marambe B, Silva GLLP, Weerahewa J, Punyawardena BVR (2012) A review

of research on homegardens in Sri Lanka. The status, importance and future perspective. Tropical

Agriculturist 160:55–125.

Reyes G, Brown S, Chapman J, Lugo AE (1992) Wood Densities of Tropical Tree Species. General

Technical Report SO-88. USDA Forest Service, Southern Forest Experiment Station, New Orleans,

Louisiana, USA.

Roshetko M, Delaney M, Hairiah K, Purnomosidhi P (2002) Carbon stocks in Indonesian homegarden

systems: Can smallholder systems be targeted for increased carbon storage?

Am J Alternative Agr 17:125–137.

Saha S, Nair PKR, Nair VD, Kumar BM (2009) Soil carbon stock in relation to plant diversity of

homegardens in Kerala, India. Agroforest Syst 76:53–65.

Schroth G, da Mota MSS, Hills T, Soto-Pinto L, Wijayanto I, Arief CW, Zepeda Y (2011) Linking

carbon, biodiversity and livelihoods near forest margins: the role of agroforestry. In: Kumar BM, Nair

PKR (eds.) Carbon sequestration potential of agroforestry systems: opportunities and challenges.

Springer, Dordrecht, pp 179–200.

Senanayake, R. L., Sangakkara, U. R., Pushpakumara, D. K. N. G. & Stamp, P. (2009). Vegetation

composition and ecological benefits of home gardens in the Meegahakiula region of Sri Lanka.

Tropical Agricultural Research, 21:1–9.

Seo S, Mendelsohn NR, Munasinghe M (2005) Climate change and agriculture in Sri Lanka: A

Ricardian valuation. Environmental Development Economics 10:581-596.

Torquebiau E, Penot E (2006) Ecology versus economics in tropical multi-strata agroforests. Tropical

Homegardens: A Time-tested Example of Sustainable Agroforestry. Eds. B.M. Kumar and P.K.R.

Nair. pp 269–282. Springer Science, Dordrecht.

Weerahewa J, Pushpakumara G, Silva P, Daulagala C, Punyawardena R, Premalal S, Miah G, Roy J et

al (2012) Are Homegarden Ecosystems Resilient to Climate Change? An Analysis of The Adaptation

Strategies of Homegardeners in Sri Lanka. APN Science Bulletin 2:22–27.

Wiersum KF (2006) Diversity and change in homegarden cultivation in Indonesia. In: Tropical

Homegardens: A Time Tested Example of Sustainable Agroforestry. Eds. B.M. Kumar and P.K.R.

Nair. pp. 13–24. Springer, the Netherlands.

Yachi S, Loreau M (1999) Biodiversity and ecosystem productivity in a fluctuating environment: the

insurance hypothesis. Ecology 96:1463–1468.

17

Supplementary material 1 List of measured tree and plant species in homegardens and their

frequency of occurrence

Local name Scientific name Family

Frequency of Occurrence

Small

(n=11)

Medium

(n=27)

Large

(n=7)

African Mahogany Kaya senagalensis Meliaceae - 1 -

Albizzia Albizia lebbeck (L.) Benth. Fabaceae - 11 -

Araliya Plumeria rubra L. Apocynaceae 2 1 2

Arecanut Areca catechu L. Arecaceae - 32 1

Ashoka Saraca asoca (Roxb.) de Wild. Fabaceae 1 - -

Avocado Persea americana Miller. Lauraceae 1 1 -

Banana Musa sp. Musaceae 41 222 55

Beli Aegle marmelos (L.) Correa Rutaceae 14 27 7

Bilin Averrhoa bilimbi L. Oxalidaceae 4 4 3

Burutha Chloroxylon swietenia DC Rutaceae 49 75 8

Cashew Anacardium occidentale L. Anacardiaceae 59 286 164

Coconut Cocos nucifera L. Arecaceae 64 211 87

Coffee Coffea arabica L. Rubiaceae 1 - -

Daminiya Grewia damine Gaertn. Tiliaceae - 1 -

Dan Syzygium caryophyllatum (L.)

Alston

Myrtaceae - 3 3

Danmon Carissa grandiflora A. DC Apocynaceae 1 1 -

Del Artocarpus incisus L. f. Moraceae 1 3 -

Delum Punica granatum L. Punicaceae 5 30 -

Diul Limonia acidissima L. Rutaceae 24 77 16

Ehela Cassia fistula L. Fabaceae - 3 -

Gliricidia Gliricidia sepium (Jacq.) Walp. Fabaceae 2 33 2

Guava Psidium guajava L. Myrtaceae 22 76 12

Halmilla Berrya cordifolia (Willd.)

Burret.

Tiliaceae - 30 2

Helamba Mitragyna tubulosa (Arn. ex

Bedd.) Kuntze

Rubiaceae 3 9 8

Hulan hik Chukrasia tabularis A. Juss Meliaceae 1 - 1

Imbul Bombax ceiba L. Bombacaceae 1 4 1

Ingini Strychnos potatorum L. f. Loganiaceae 1 1 -

Ipil Ipil Leucaena leucocephala (Lam.)

de Wit

Fabaceae 9 34 4

Jam gaha Muntingia calabura L. Tiliaceae 1 6 1

Jambu Syzygium jambos (L.) Alston Myrtaceae 2 4 1

Kaduru Cerbera odollam Gaertn. Apocynaceae - 11 -

Kamaranga Averrhoa carambola L. Oxalidaceae - 1 -

Kara Canthium coromandelicum

(Burm. f.) Alston

Rubiaceae 1 1 -

Karapincha Murraya koenigii (L.) Spreng Rutaceae 2 6 4

Katakala Bridelia retusa (L.) Spreng. Euphorbiaceae - 1 -

Katu Anoda Annona muricata L. Annonaceae - 5 -

Katurumurunga Sesbania grandiflora (L.) Poir. Fabaceae - 6 -

18

Kon Schleichera oleosa (Lour.) Oken Sapindaceae 3 4 1

Komuk Terminalia arjuna (Roxb.)

Wight & Arn.

Combretaceae - 1 -

Kos Artocarpus heterophyllus Lam. Moraceae 20 53 10

Kottamba Terminalia catappa L. Combretaceae - 1 1

Lime Citrus aurantifolia (Christm. &

Panzer) Swingle

Rutaceae 17 17 10

Liyan Homalium zeylanicum (Gardn.)

Benth.

Flacourtiaceae - 1 -

Lunumidella Melia azedarach L. Meliaceae - 1 -

Maila Bauhinia racemosa Lam. Fabaceae 2 40 9

Malitha Woodfordia fruticos (L.) Kurz Lythraceae - - 1

Mango Mangifera indica L. Anacardiaceae 38 149 31

Mara Samanea saman (Jacq.) Merr. Fabaceae - 1 -

Marthondi Lawsonia inermis L. Lythraceae - 149 -

May-mara Delonix regia (Bojer ex Hook.)

Raf.

Fabaceae - 2 2

Munamal Mimusops elengi Sapotaceae 1 5 6

Murunga Moringa oleifera Lam. Moringaceae 1 33 7

Neem Azadirechta indica A. Juss. Meliaceae 200 648 164

Orange Citrus sinensis L. Rutaceae 10 16 2

Palu Manilkara hexandra (Roxb.)

Dubard

Sapotaceae - 4 -

Papaya Carica papaya L. Caricaceae 7 52 20

Sepalika Nyctanthes arbor-tristis L. Nyctanthaceae 6 2 -

Sal Couroupita surinamensis Lecythidaceae - 1 -

Sidaran Citrus medica L. Rutaceae - 1 -

Siyambala Tamarindus indica L. Fabaceae 15 36 11

Suria mara Albizia lebbeck Fabaceae 7 17 3

Turpentine Eucalyptus camaldulensis

Dehnhartd

Myrtaceae 5 5 -

Thekka Tectona grandis L.f. Verbanaceae 60 154 23

Uguressa Flacourtia indica (Burm. f.)

Merr.

Flacourtiaceae 2 4 -

Wa Cassia siamea Fabaceae 2 11 -

Wal Ambarella Spondias pinnata (L. f.) Kurz Anacardiaceae 6 4 3

Welianoda Annona reticulata L. Annonaceae 68 207 26

Weralu Elaeocarpus serratus L. Elaeocarpaceae 8 7 3

Number of non-

identified species

5 7 2