Analysis of the Impact of Built Environment on Coastline Ecosystem Services and Values

The degradation of any ecosystem services (ES) and the benefits human being enjoy from nature freely involve multifaceted processes such as those in built environment. The ecological security and multiple functions of the coastal zone of Dar es Salaam is of paramount importance to the sustainability of its natural and anthropogenic systems. Therefore, permanent/temporary conversion of a piece of land for construction space provisions profoundly affects the functionality and connectedness of nature–anthropogenic ecosystem. This study quantified land use landcover changes from Landsat satellite imageries, then evaluated the changes using recognizable coefficients for ecosystem services values (ESV). Applying Geographical Information System (GIS) techniques, the impacts of built environment on ESV were analysed using land use landcover change (LULCC) transfer matrix, carbon stock dynamics and soil erosion influence on soil fauna. The results revealed expansive BE from 10.6 percent in 1995 to 22.8 percent in 2016. Loss of forest by 64.5 percent in the study period explained the declining total ESV by 56.1 percent and per capita recreation potential by 2.3 percent. Similarly, decreasing forest cover led to high carbon dioxide emission, notably, the 353.24 t CO2 ha-1 yr-1 in the period between 1995 and 2005. Furthermore, in 1995 bushland experienced high soil erosion while in 2016 built environment displayed a similar trend as the rest of the land use landcover (LULC) classes. From geospatial analysis, the southern area displayed significant vegetation cover change as compared to the built environment dominant in the northern section of Dar es Salaam coastline. Initiative to reducing built environment by 1 percent saves forest loss by 5.28 percent and carbon sequestration at a tune of 28.95 t CO2 ha-1 yr-1; hence improves ecological services values by 4.60 percent.


INTRODUCTION
The features, structures and facilities made, arranged, maintained and controlled by humans are rising globally (Kane & Clark, 2019;Lobora et al., 2017;Seto et al., 2012). These urban places also referred to as built environment (BE) connote manmade spaces and places created or modified for various uses and users (Shahraki, 2017). Studies by Pinho and Lúcio (2017) reveal that the driving force on human thoughts, actions and experiences on BE is space value provision for recreation, workplaces and residence. According to Murdoch et al. (2018), burgeoning human population across global regions pressurizes the delivery of such space values. While BE assure the provision of exchange values, use values, image values, social values, cultural values and environmental values (Khademagha et al., 2016); the intended/unintended, direct/indirect, positive/negative both short and long-term impacts from LULCC of BE is a colossal challenge of our time (Aerts et al., 2014;Carmona, 2019). One of the greatest consequences of urban development is LULCC such as loss of vegetation cover, agricultural land and natural resource base meanwhile influencing environmental degradation and pollution (Halmy, Gessler, Hicke, & Salem, 2015;Karki et al., 2018).
The nature benefits to human wellbeing is of paramount importance (van Heezik & Brymer, 2018). Researchers Elmqvist et al. (2015) epitomize that presence of biodiversity in urban landscapes ensure humans with the availability of essential ES such as food provisioning, pollution reduction, climate regulation, carbon sequestration, and nutrient cycling. According to Butt et al. (2014) and Parris et al. (2018), BEs take, consume and sit permanently on a piece of land leading to landscape fragments and loss of ecological habitats. Such absence of ecological habitats connectedness reduces the landscape capacity to produce ES as well as access to ES by the ecological community (Cordingley et al., 2015). Aggressive construction activities and vegetation clearance cause soil erosion (Ligonja & Shrestha, 2015), a key contributor to siltation into hydrological systems. Therefore, in high soil erosion risk zone (Panagos et al., 2012) like Dar es Salaam coast could be one among the drivers for hazardous, notably repeatedly flooding along Dar es Salaam coastline. Siltation and sedimentation in the Dar es Salaam Metropolitan make this coastal lowland and high soil erosion risk zone to face more hazardous events (Sakijege et al. 2012).
While microfauna is responsible for soil pores, loss of vegetation cover influence soil erosion events; such occurrences displace microfauna with negative effects on soil ecological system (Orgiazzi & Panagos 2018;Shanshan et al., 2018). Study by Gashaw et al. (2014) content that loss of soil microfauna and vegetation cover impacts carbon cycle and stocks thus primary productivity. According to Senga (2014) and Zari (2014), landscape physical degradation including soil erosion disrupts the hydrological system and net biome productivity. Feedback loops on the aforementioned variables reduce nature resilience henceforth declining ES (Keller et al., 2018).
Recently, recognition, quantification and performance of ES in Anthropocene epoch has grown tremendously (Thomson and Newman, 2018;Zari, 2019), albeit ongoing global debate on urban biodiversity and its involvement in ES production and supply (Gómez-Baggethun & Barton, 2013;Hurley & Emery, 2018). Earliest work on ES studies by Costanza et al. (1997) set a stage on ES valuation proxies; today, literature reveals three approaches to ES valuation; willingness to pay (WTP), LULCC based methods and ecological process-benefit assessments (Yuan et al., 2019). According to Hepelwa (2013) and Schmidt et al. (2016) technological advancement in remote sensing (RS) and geographical information system (GIS) has efficiently improved LULCC based evaluation methods as compared to the others. Similarly, according to Xie et al. (2017), development of LULCC proxies for ES valuation delivered a useful approach to evaluating the impact of LULCC in various landscapes including urban centres. Lal (2015); Adhikari and Hartemink (2016) contend that GIS provides a platform to track and integrate ES, carbon stock and soil erosion data thus the supply of quantified statistics enable informed decision on green and inclusive cities, in particular, SDG 11. Despite this, limited scholarly work has devoted to explore BE spatio-temporal changes and its effect on ES, notably along Dar es Salaam coastline. Therefore, this study investigated the pattern of LULCC along Dar es Salaam coastline, with a focus on identifying evidence for urban ES changes. Application of geospatial approaches (Gombe et al. 2017;Mkalawa 2016) for monitoring LULCC using Landsat imageries enabled analysis of the built-up area, carbon stocks and soil erosion; hence quantifying impacts of such changes to soil fauna and ES values along Dar es Salaam coastline in Tanzania

Study Area
Dar es Salaam lies between Latitude 6° 3' 43.09" to 7° 10' 47.35" S and Longitude 39° 6' 36.37" to 39° 33' 5.66" E at 24 meters above sea level. The Metropolitan size is approximately 1800 km 2 at 3:1 land: water coverage ratio along the southwestern coast of the Indian Ocean. This study concentrated on a section of 714.37 km 2 , which is 53 percent of the landmass of Dar es Salaam metropolis; the maximum inland distance was 16.6 km while the minimum was 3.3 km. Dar es Salaam experience hot-humid climate greatly influenced by the northeast and southeast monsoon. The Metropolis receives an average annual rainfall of 172 mm, has an average annual temperature of 29 o C and humidity record of 96 percent in the morning and 67 percent in the afternoon. The coastal shrubs, miombo woodland, coastal swamps and mangrove trees represent the main natural vegetation cover type along the coastline. Governed in five districts of Kigamboni, Temeke, Ilala, Kinondoni and Ubungo. Dar es Salaam is the largest and most industrialized Metropolis in Tanzania. Its population has grown from 843,090 in 1978 to 5,465,420 in 2016 with 94 percent being urbanites (Gombe et al. 2017;Mkalawa 2016). This study concentrated on the coastline of Dar es Salaam Metropolitan ( Figure 1)

Data Sources, LULC Classes and Accuracy Assessment
Landsat Thematic Mapper (TM) images for June 1995 and 2005; and Landsat Operational Land Imager (OLI) for July 2016 were freely accessed and downloaded from United States Geological Survey (USGS) Earth Explorer (EE). Landsat scenes used were from path 166 row 65 for Dar es Salaam, Tanzania, with 30m pixel resolutions and minimal cloud cover (<10 percent). A series of processes involving data acquisition, data preprocessing, unsupervised and supervised classification, and post-classification comparison provided useful data. On the other hand, GPS based field survey supplied inputs for ground-truthing and accuracy assessment.
According to Iqbal and Khan 92014) as well as Al-Bilbisi (2019), simplicity of Landsat imagery interpretation depends on image pre-processing activities. In this study, image pre-processing underwent image registration and projection from UTM WGS 184_Zone_37N its downloadable format into UTM WGS 184_Zone_37S, the exact location of the study area. Thereafter, removal of top_of_atmosphere (TOA) reflectance activity and thereafter image enhancement procedure. Classification involved both unsupervised and supervised techniques. In unsupervised classification the image's final product is on the basis of LULC properties; execution of supervised classification using 35 training samples with corresponding spectral signatures (Awad et al. 2019;Diesing et al. 2016) as inputs to maximum likelihood algorithm in ArcGIS 10.3 provided seven LULC classes (Table 1) (URT, 2010). The methodological flow chart in ArcMap 10.3 is presented in Figure 2.

LULC Classes National Land Cover Description
Forest An area of land with at least 0.5 ha, with a minimum tree crown cover of 10 percent with tree reached a minimum height of 3m or above at maturity in situ. It includes montane, lowland, mangrove and plantation forests, woodlands and thickets. The LULC classes obtained underwent accuracy assessment (Rwanga & Ndambuki, 2017). According to Susilo et al. (2018), any mismatch between reference data and image LULC classes indicates poor image classification. In estimating accuracy assessment, the study used the very recent image (2016) and 34 ground-truth reference data collected by global positioning system (GPS) tool (Rwanga and Ndambuki 2017;Susilo et al. 2018).

Soil Erosion Estimation
Understanding and quantification of detaching soil particles (soil erosion) and their relocation or displacement due to physical factors (Nyawade et al., 2019) is important in defining and setting strategies for resources conservation and management. According to Yesuph and Dagnew (2019), an average soil erosion in tons per hectare per year (t ha -1 yr-1) depends on rainfall erosivity, soil erodibility, cover management, length and slope steepness and supporting practices. This study, using cross-tabulation analysis of RUSLE Model (Equation 1) and LULC changes polygon in ArcGIS v10.3 combined all the aforementioned factors to quantified soil erosion in each LULC classes (Koirala et al., 2019).

Loss of Soil Fauna
Globally, the effect of urbanization on the natural landscape is noticeable (You & Yang, 2017).

Carbon Stock Estimation
Studies by Bindu et al. (2018) and Ekoungoulou et al. (2015) clearly documents the carbon stock dynamics with respect to LULC changes. The interlacing relationship between rapidly expanding BE and climate change scenarios calls for concrete data on LULCcarbon interactions (Mannan et al., 2019). In view of this, the study estimated AGC from AGB and BGC from BGB. Then applying NAFORMA (Malimbwi & Zahabu, 2014), developed carbon proxy equivalence as 59.5 t/ha for AGB and 18.2 t/ha for BGB, the relation between AGB and BGB is provided by Equation 2. Conversion of GB into GC is by default factor (Equation 3). Recent estimate (Alavaisha & Mangora, 2016) of mangrove-based carbon stocks along Dar es Salaam coastline provide an equivalence of 26.18 t C/ha. The remote sensing and GIS extracted LULC changes data of the study area providing useful input for the computation of carbon stocks dynamics along the coastline.

Ecosystem Services Values Function (ESVf) Estimation
ESVf measures the capacity of an ecosystem to produce and supply ES to landscape habitats (Rieb et al., 2017); and that its understanding is a vital requirement across a range of stakeholders with fatal decision in green cities, sustainability and climate changes initiatives (Bidegain et al., 2019). That being the case, the study extracted the quantity of each LULC classes converted into built environment, hence treated them as lost ES; thereafter their monetary values were obtained through the application of ES values equivalences from (Costanza et al., 1997). Despite several existing methods for ES valuation (Neugarten et al., 2018), the ESVf in Table 3, which is an extract from (Costanza et al., 1997) and (Yuan et al., 2019) has been instrumental in understanding the cost of LULC conversion. Equation 4 provides both the relation between ESVf and LULC data as well as the approach to compute the monetary value of the lost ES from 1995 to 2016; = ∑ × …………………… (4) Where: is the ecosystem service values function; is the LULC area size in hectares; is the coefficient of ecosystem service function (US $ ha-1 yr-1) for LULC class k.

LULC Change Dynamics
Recently, the use of LULC change approach in monitoring and evaluating the impact of anthropogenic activities from Landsat satellite imageries have proved to provide useful environmental data, particularly, on the past time (Kukulska-Kozieł et al., 2019). The ArcGIS unsupervised and supervised Landsat classification procedures (Figure 2) provided the results in Figure 3, showing that agriculture and built-up landscape increased significantly between 1995 to 2016; while vegetation cover, particularly in the southern parts of the coastline (Figure 4), experienced a declining trend. In the same period, the annual rate of change in grassland and forest cover displayed a high negative value than the rest of the LULC classes (Table 4). Such value indicates loss of green spaces along Dar es Salaam coastline metropolitan, which according to (Mkalawa, 2016;Mkalawa & Haixiao, 2014) increasing built environment and decreasing vegetation cover revealed is linked to the coincidence of population growth, economic growth and demand for infrastructures in Dar es Salaam. It is important to observe recommended plot coverage and roadside strips if green cover is to be maintained in the human being dominated environment such as Dar es Salaam. As a product of urbanization, BE promotes various massive infrastructure development to meet the demands of the growing number of urbanites. This expanding BEs take, consume and seat permanently on a piece of land, thus loss of habitats and ecological species due to fragmentation, isolation and extinction (Aerts et al., 2014;Zheng et al., 2017). Decreasing ecological species affect the ecosystem components, functions and structure hence qualities and quantities of ESS produced in respective landscapes. Such reduced ESS activates more resources extraction beyond the landscape's boundaries; a situation that leads to further impacts associated with expanding agricultural landscape and BE (Table 4 & Table 5). The established requirement for food production to feed the urbanites, energy (firewood and charcoal) (Malimbwi & Zahabu, 2014;URT, 2010) and raw materials for infrastructure development, are the driving forces for such expansion.  Table 5). The reciprocal relationship between increased population and decreasing vegetation cover ( Figure 5) displayed decreasing per capita recreation potential. Statistically, the urbanites level of enjoyment of vegetation-based ES appear to be decreasing considerably (R 2 = 0.94). The impact of poor recreation and aesthetic potentials in the anthropogenic environment is associated with increasing low social engagement, ineffective physical activities and low reasoning capacity and fear, notably in childhood (de Sa & Ardern, 2014).

Estimation of Carbon Stock
Analysis of LULC carbon stock using (Malimbwi & Zahabu, 2014;URT, 2010) carbon estimation equivalence, displayed a declining trend to both above ground carbon (AGC) and below ground carbon (BGC) (R 2 = 0.82), for forest covers ( Figure  6). Applying a factor of 3.6667 in relating carbon and carbon dioxide, the study quantified the amount of carbon dioxide emission ( Table 6). Decreasing vegetation cover interferes carbon cycle and other biogeochemical cycles (Butt et al., 2014;Parris et al., 2018), leading to carbon dioxide in the biosphere; continual rising of this greenhouse gases (GHGs) ( Table 6) contributes to climate change incidences at Dar es Salaam coastline. Furthermore, expansion of BE linked impervious surfaces disrupts hydrological cycles hence local microclimate dynamics. According to (Parris, 2016;Watkiss et al., 2011) climate change impacts is profoundly leading to increased flooding events, Dar es Salaam every rain season risk events.  From Figure 7, the analysis provides that in 1995, bushland had the highest soil erosion while in 2016 agriculture and built-up landscape displayed highest soil erosion. Studies show that expanding BE increases impervious surface (Baker et al. 2019), a blockage structure between lithosphere, atmosphere, (Costanza et al., 1997) hydrosphere and geosphere thus destroying soil porosity, structure, texture, density as well as functions of biogeochemical cycles. Limited porosity, water and soil aeration flow in BE influence soil chemical and physical properties hence ES production and supply dynamics. Meanwhile, expanding BE insinuates huge urban-runoff abetting soil erosion, biodiversity loss and pollution of hydrological resources. The three aforementioned factors contribute significantly to poor quality and quantity, and or loss of ES in the respective landscape. High soil erosion in 2016 as compared to 1995 is explained by various construction activities in Dar es Salaam central business district (DCBD). In this period, the DCBD experienced commercial buildings skyrocketing and urban road infrastructures, notably, Dar es Salaam Bus Rapid Transit (DART) project (Rizzo 2014;de Sa and Ardern 2014). Therefore, extraction of construction materials as well as urban-runoff is likely to be the cause of such dynamics in soil erosion quantities.

Soil Fauna Lost/Displaced by Soil Erosion
Quantification of soil fauna displaced/lost by soil erosion (Table 7) involved the use of soil fauna coefficients (Orgiazzi & Panagos, 2018). Results show that soil fauna loss increased significantly in Nematodes (21.7 percent), Earthworms (18.2 percent) and 17.9 percent in Mites, Collembola and Enchytraeids respectively. The diverse and numerous plant and soil microfauna are central to soil ecosystem natural settings, in which the microbial activities are responsible for soil physical, chemical and biological properties (Jacoby et al., 2017). Their decomposition and nitrogen-fixing characteristics account significantly for energy and nutrient circling. Thus, loss/displacement of soil fauna due to BE processes, activities and expansion is indirectly or directly interfering soil formation and properties. In the context of ES production and supply, loss/displacement of soil fauna by soil erosion interferes with the functioning of biogeochemical cycles, water infiltration and aeration, soil formation and properties and atmospheric gas regulations. Studies (Lal, 2014(Lal, , 2015Orgiazzi & Panagos, 2018) epitomizes that the interaction between soil biodiversity and soil erosion are complex interlinked and bidirectional and that the magnitude of negative effect to each other lies in the presence of anthropogenic issue.   Table 8). During the study period, the value of ESVf changed too (Figure 8 (a) -(d)). Analysis of ESVf on LULC converted into BE displayed significant changes in the study area (Table 9).  In a nutshell, the relationship between soil, BE and biodiversity is complex, and as the world ecosystem face more urbanization; urban centres are likely to become central sources of ES and values. Attentiveness on biodiversity offset in biodiversity-infrastructures development nexus improves integration of biodiversity habitat strips in the anthropogenic dominated environment. Similarly understanding the intertwined complex interactions between soil ecology and soil erosion maximizes ES production and supply, thru sustainable land use planning.