Hotspots of land use/land cover change around Bolgoda wetland, Sri Lanka

* Corresponding author (chamanthaathapaththu@gmail.com; https://orcid.org/0000-0002-4808-1474) This article is published under the Creative Commons CC-BY-ND License (http://creativecommons.org/licenses/by-nd/4.0/). This license permits use, distribution and reproduction, commercial and non-commercial, provided that the original work is properly cited and is not changed in anyway. Abstract: Wetlands are among the most productive ecosystems in the world. Urban wetlands are increasingly aff ected by population growth and developmental activities. A buff er region of 60 m from the Bolgoda lake boundary was gazetted as an Environmental Protection Area (EPA) by the Central Environmental Authority in 2009 as it is the largest freshwater wetland in Sri Lanka. This study attempts to quantify the land use changes during the period of 2001 to 2019 using GIS and to identify the ‘hot spots’ where a signifi cant change in land use occurred. Land use/land cover (LULC) changed in the Bolgoda wetland area disproportionately. The greatest proportion of loss of natural area was observed in dense vegetation, where a 63.35 % decrease was apparent from 2000 to 2019. In contrast, the residential and commercial areas increased and in 2019 the change reached 49.62 % and 68.57 %, respectively. Hotspots were identifi ed at Grama Niladhari divisions that belong to Kesbewa and Bandaragama DS divisions with the largest change in LULC. Thus, the results of the investigation provide vital information for the conservation and sustainable use of wetland resources in a rapidly expanding urban landscape.


INTRODUCTION
Wetlands are complex and highly productive ecosystems with high ecological, social, and economic values. Many wetlands are high in biodiversity and provide habitats for fl ora and fauna of signifi cant ecological importance (Jinadasa et al., 1992;Punchihewa et al., 2017). These habitats are particularly important for livelihood of the communities who live in the area. For instance, fi sheries and tourism are dependent on wetlands (Gachhadar et al., 2004). Ecosystem services provided by wetlands include absorption of pollutants, carbon sequestration, groundwater recharge, and disaster mitigation (Chmura et al., 2003;Chen et al., 2008;Ramsar Convention Secretariat, 2014). Yet human induced land use/land cover (LULC) changes aff ect the wetlands worldwide as well as services provided by them (Zhao et al., 2004;Zorrilla-Miras et al., 2014). Tooth et al. (2015) state that the wetlands provide major ecosystem services, including water and food supply, which however occur in climatically varying and moisture stressed environments that can be treated as 'hotspots.' Therefore, monitoring land use changes around wetlands is important in the conservation and management of wetlands.
Remote sensing (RS) and Geographical Information System (GIS) are increasingly being used to investigate land use changes (Rogan & Chen, 2004;Treitz & Rogan, 2004). RS and GIS are considered appropriate for wetland monitoring, especially when the extent of land considered is large. Moreover, the availability of multitemporal images allows temporal analysis of land use changes over time (Ozesmi & Bauer, 2002).
Bolgoda wetland has gained attention of policy makers, since it is the largest fresh water wetland in

September 2020
Journal of the National Science Foundation of Sri Lanka 48 (3) Sri Lanka with high ecological, social and economic value (Silva et al., 2013). Located in the Western Province it comprises north and south lakes. In 2009, the Bolgoda lake and a buff er region of 60 meters from the lake boundary were gazetted as an Environmental Protection area (EPA) under the Gazette Notifi cation No. 1634/23 (Central Environmental Authority, 2018. Nevertheless, it has deteriorated in both quantity and quality due to unplanned developmental activities, growth of population, and unsustainable urbanisation (Dahanayaka et al., 2016).
Monitoring LULC changes presents important insights for management of natural habitats and helps decision making in controlling the balance between natural and human altered landscapes (Giner & Rogan, 2012). In this context, the present study was carried out with two objectives: to quantify the land use changes during the period 2001 to 2019; and to identify the hotspots of LULC change where signifi cant changes in land uses were apparent. The results of the investigation provide vital information for the conservation and sustainable use of wetland resources in rapidly expanding urban landscapes.

Study region
The Bolgoda lake (6°52´ and 6°39´ North latitudes and 79°52´ and 80°0´ East longitudes) and a buff er region of 2 km from the lake boundary were considered as the study region with a total area of 140 km 2 (Figure 1). The study region is located under the divisional secretariat (DS) divisions of Ratmalana, Kesbewa, Moratuwa, Panadura, Bandaragama, and Kalutara.

Data collection
Landsat images corrected for surface refl ectance were obtained from the United States Geological Survey (USGS) (Masek et al., 2006). These particular images were selected based on image availability and the presence of less than 20 % land cloud cover (Table 1).
Moreover, a rapid assessment to investigate fi eld conditions was carried out by visiting the study area in 2018 followed by semi-structured interviews with 25 residents of the area to understand the underlying reasons for land use changes.

Image classifi cation
Satellite images from Landsat 5 and Landsat 8 were used in LULC classifi cation. Spectral bands of images were selected based on corresponding wavelengths of TM and OLI sensors. Six spectral bands were used in the classifi cation: blue, green, red, NIR (Near Infrared), SWIR (Short Wave Infrared) 1, and SWIR 2.
Through fi eld observations and visual interpretations, six LULC types were identifi ed: water body, commercial areas, residential areas, dense vegetation, sparse vegetation and bare land (Table 2).
Satellite images were classifi ed using the supervised classifi cation method in ArcMap version 10.5. As the fi rst step, pixels with known land use categories were selected to prepare the training sample. These training sites for each LULC were selected based on high resolution images from Google Earth. Historical images from Google Earth were used to identify land use characteristics of the past. NDVI (Normalied Diff erence Vegetation Index) maps were prepared to identify the vegetation condition of each year. Training sites for vegetation classifi cation were selected based on NDVI values. Due to the lack of high-resolution images, the same training pixels from 2005 were used for classifi cation of the 2001 image.
The pixels of the images were then assigned to each LULC category according to statistical similarity to the training site's pixels. Classifi cation was done based on Maximum Likelihood Classifi er Algorithm.

Accuracy assessment
The accuracy of the classifi ed images was determined using the confusion matrix method. The classifi cation results were compared with testing site pixels. These testing sites were obtained using high resolution images from Google Earth. For each LULC type, 60 reference points were used. Accuracy assessment was not done for 2001 due to lack of high-resolution images.
Overall accuracies of all the classifi ed images were above 75 % and the Kappa coeffi cients were above 0.7 as shown in Table 3. These values show the statistical agreement between the classifi ed image and the reference data (Congalton & Green, 2002).

Identifying hotspots of land use/land cover change
To identify hotspots of LULC change from vegetation to build up region, both residential and commercial regions from 2001 and 2019 were extracted as separate layers. Then both layers were combined using the union tool in ArcMap. From the combined layer, new built up areas which appeared in 2019 were extracted again as a separate layer.

Results
The area of six land use types in each year diff ered between the years (Table 4). Bare lands and water bodies showed relatively low percentage area change (Figures 2  and 3). Commercial land use has an increase of 68.57 %. Furthermore, residential land use areas have shown a 49.62 % increase while dense vegetation has decreased by 63.35 %. Sparse vegetation, on the other hand, was increased by 12.51 %. These results indicate that dense vegetation has transformed to sparse vegetation and in some parts, to residential and commercial land use. As a result, residential areas have become the predominant land use type and sparse vegetation has become the major vegetation type around Bolgoda lake.
A rapid fi eld assessment revealed that utilisation of the wetland by the residents were not prominent; livestock grazing, small scale collection of leafy vegetables and freshwater fi shing were the only signifi cant signs of exploitation of wetland resources. Paddy lands in the area have been abandoned due to non-availability of labour. At present the Bolgoda wetland area has become a thriving hub for tourism and recreational activities including bird watching and water sports such as rowing. Thus, the local community may have moved to alternative livelihoods such as tourism with the increasing potential of hospitality activities in semi-urban areas targeting both local and international visitors (personal observations). Over time, most of these paddy lands have gradually been converted to wetlands. All respondents to interviews indicated the need for conserving the wetland to provide aesthetic value, clean neighbourhood and livelihood opportunities especially in hospitality industry.

Discussion
With rapid urban development, wetlands and surrounding landscapes in many parts of the world have faced changes in LULC ( Obiefuna et al., 2012;Zhang et al., 2015). For instance, various urban wetlands have undergone changes due to infrastructural development including construction of houses and buildings (Sithole & Goredema, 2013).The situation is similar around Bolgoda lake in Sri Lanka.
As expected, LULC has changed, although disproportionately, in the Bolgoda wetland area. The greatest proportion of loss of natural area was observed in dense vegetation, where a 63.35 % decrease was apparent from 2000 to 2019. Dense vegetation has been replaced by the residential and commercial areas that have increased by 49.62 % and 68.57 %, respectively. Most of the changes have occurred during the ten year period from 2005 to 2015. Studies report that during the same period urban expansion has increased by more than 30 % when compared to the period before 1985, due to population growth and increase in commercial and industrial activities in the area (Weerakoon, 2017). Another reason is that following the end of civil confl ict in 2009, many economic opportunities were opened up, which facilitated growth of urban activities. As a result, the wetlands, cultivations and even home gardens have been transformed into infrastructural and commercial areas. Field observations confi rmed that houses and commercial buildings including hotels and cottage industries are encroaching the natural environment. Most of such lands were owned by private owners and high land prices have encouraged people to convert their lands into other uses.
Several factors have contributed to the increase in sparse vegetation. Abandoned paddy fi elds have been converted to marshes over-time. On the other hand, the amount of water logging has increased in the area resulting frequent fl oods (Jayasinghe & Rajapakse, 2017). Thus, sparse vegetation has gradually increased, which is refl ected in low NDVI values in the aerial maps. Another major fi nding of this study is the hotspots: Kesbewa and Bandaragama DS divisions have undergone the largest change in LULC. The reason behind this is linked to the population increase where these DS divisions showed the highest rate of population growth according to the population census in 2001 and 2012 (Department of Census and Statistics Sri Lanka, 2019). Thus, high rate of population increase may be the underlying reason for land use conversion in these regions.
Changes of natural landscapes result in many issues. As a consequence of LULC change, property values of land could be aff ected (Gwamna & Yusoff , 2016). On the other hand, the LULC is considered as one of the key driving factors of global change which refl ects the anthropogenic impacts on the environment (Dewan et al., 2012). Understanding the degree of such changes over time is crucial for urban planning as well as conservation of natural habitats (Mirkatouli et al., 2015).
For instance, LULC changes from natural ecosystems to other land uses aff ect ecosystem services provided by wetlands; including hydrological functions (Zhan et al., 2019), carbon sequestration capacity (Xu et al., 2018); impairment of water quality (Houlahan & Findlay, 2004), loss of green spaces, and reduction of biodiversity. Although the LULC changes are visible, their long term and irreversible impacts often go unnoticed. Biodiversity worldwide is reported to be aff ected by human-induced changes in the structure of ecosystems (Sinha & Sharma, 2006). LULC change is regarded as the primary cause of global biodiversity crisis, especially due to habitat loss (Jackson & Sax, 2010). Moreover, LULC is the key culprit for global species extinction (Didham et al., 2012). For instance, Sri Lankan wetlands with increased pollution levels have negatively impacted amphibian fauna (Jayawardena et al., 2013, Priyadarshani et al., 2015 which is most vulnerable to environmental contaminants and have been reported to show a decline globally (Gallant et al., 2007).
To achieve the Sustainable Development Goals (SDGs) by 2030, healthy wetlands are essential as they provide multiple services to the local community. Goal number 6 (clean water and sanitation), goal number 11 (sustainable cities and communities), and goal number 14 (life on land) are directly related to urban wetlands. In the Colombo District, urbanisation has been identifi ed as the main driver of wetland modifi cation (Hettiarachchi

Journal of the National Science Foundation of Sri Lanka 48(3)
September 2020 et al., 2014). Especially in the Bolgoda area, increased commercialisation has been reported in previous studies (Piyadasa & Chandrasekara, 2010). In case of Bolgoda, 'hotspots of LULC change' where the ecosystem could continue to degrade, losing its potential to contribute positively to environmental health, warrant special attention. The city of Colombo has been declared as one of the fi rst wetland cities in the world (Ramsar Convention Secretariat, 2018). In this context, proper conservation management plans for Bolgoda wetland is crucial in securing ecosystem services off ered by this wetland.

CONCLUSION
The decrease of dense vegetation with an increase in residential and commercial areas was evident from 2001 to 2019 due to transformation of natural land to urban areas. Kesbewa and Bandaragama DS divisions demonstrate the highest LULC change and can be regarded as 'LULC hotspots'. As these changes could result in various environmental and socio-economic issues in future, appropriate urban planning as well as implementation of policies and regulations are essential to ensure the balance between nature and development.