Translator Disclaimer
7 April 2021 Environmental Factors and Spatial Heterogeneity Affect Occupancy Estimates of Waterbirds in Peninsular Malaysia
Author Affiliations +
Abstract

In Malaysia, multiple land use by humans has caused substantial losses of wetland ecosystems, and shrinkage of the populations, habitat and food bases of avian species. However, studies of avian populations, especially of waterbirds, is important, allowing us to understand the complexity of the wetland ecosystem structure, and also develop appropriate management techniques with robust monitoring tools to ensure the ecological sustainability of wetlands. This study aimed to determine the eco-climatic factors influencing the occurrence of waterbirds and to develop habitat suitability models for thePaya Indah (PIW) and Putrajaya wetlands (PW), Malaysia. A distance sampling point count technique using stratified random design was employed to survey the wetlands from November 2016 to January 2019. A total of 57 sampling points at 14 lakes at PIW and 54 sampling points at 24 lakes at PW were chosen. An automatic linear modelling algorithm and geographic information systems were employed to compute the importance ratios of 17 eco-climatic factors (hydrology 9; climate 5; waterscape 1 and landscape 2). The results revealed that all individual and estimated indices for observed waterbirds were significant. The automatic linear modelling algorithm results for PIW waterbirds also showed that the maximum and minimum weights of the factors were land cover and water dissolved oxygen, while in PW they were atmospheric pressure and Normalized Difference Water Index (NDWI). The maximum and minimum weights of the factors for waterbirds in PIW were water turbidity and electrical conductivity, while at PW they were atmospheric pressure and six water parameters. Large areas of Putrajaya Wetland were classified as more suitable for waterbirds than Paya Indah Wetland due to the favourable water pH, atmospheric pressure and land cover (forage availability). Thus, the models' adoption as a management tool can help in the maintenance of the wetlands' habitat quality and management effectiveness of waterbird species.

The study of waterbird populations is important for understanding the complexity of wetland ecosystems and for providing information on various types of wetlands in an ecosystem (Weller 1999; Matsinos & Wolf 2003; Burger & Eichhorst 2005; Amat & Green 2010; Rahman & Ismail 2018). In Malaysia, of a total of 815 species of birds recorded at wetlands, 170 are waterbirds (Zakaria & Rajpar 2013; Malaysian Nature Society Bird Conservation Council 2015). Some waterbird species occur throughout the country within one of the most varied and diverse wetland ecosystems mainly due to the heterogeneity of vegetation phytosociological characteristics, unpredictable rainfall patterns and occurrence of different contiguous landscapes (Rajpar & Zakaria 2014). Unfortunately, wetland ecosystems are threatened with extinction due to habitat loss, invasive species and human intervention (World Bank Report 2011). Urbanisation and increased agricultural activities, such as the development of oil palm plantations, have replaced large areas of wetland vegetation (Mundava et al. 2012). Despite the undue human pressure on Malaysian wetlands, waterbird density and distribution are important for vegetation dynamics, availability of water and food resources, protection from predators and climatic condition (Zakaria & Rajpar 2010; Rajpar & Zakaria 2013). Inland wetlands most especially in highly urbanized areas are critical to waterbirds conservation (Morganti et al. 2019). Paya Indah Wetland (PIW) and Putrajaya Wetland (PW) are both highly urbanized while also the largest wetlands in Selangor State (the most populous and developed state in Malaysia) and in Putrajaya (the fastest growing region) (Ho 2006). These two homogenous wetlands may, due to their proximity (approximately 10 km) to different wetland types (natural and artificial), support varied abundances and distributions of waterbirds species in relation to their ecological and micro-climatic factors.

Past studies have revealed both direct and indirect relationships between waterbird distribution and the environmental factors affecting wetlands, such as rainfall, atmospheric temperature, relative humidity, water temperature, land cover (also including use), wind speed, barometric pressure, flood level, tree species and height, food types, and types of water body (Ludwig et al. 2006; Ajonina et al. 2009; Fasola et al. 2009; Ismail & Rahman 2013; Zainul-Abidin et al. 2017; Morganti et al. 2019). Because of the drastic reduction in waterbird populations as well as shrinkage in wetlands globally (Wainger & Mazzotta 2011; Gumbricht et al. 2017; Wetland International 2018) it is very important to establish the distribution areas of waterbirds with immense scientific value. Fluctuations in the distribution of waterbirds reflects the spatial variability of environmental factors on micro and macro scales (Fasola et al. 2009), while spatiotemporal variation in bird assemblages is associated with seasonal variations in ecological and environmental variables (Paolini et al. 2018; Santillan et al. 2018).

Habitat suitability models (HSMs) are an effective tool for resolving the fundamental research problems of determining the ecological and microclimatic factors that significantly influence waterbird distribution within homogenous wetlands in urbanized areas. Since the 1990s, the use of HSMs has facilitated the prediction of species occurrence (Hirzel & Lay 2008), presence or absence (Rushton et al. 2004) or distribution (Franklin 1995; Guisan& Zimmermann 2000) in relation to environmental variables in a particular area. Different approaches have been employed for selecting variables or weighting factors during habitat suitability modelling. These approaches include generalised additive models (Thuiller 2003; Thuiller et al. 2009), generalised linear models (Thuiller 2003; Thuiller et al. 2009), generalised regression analysis and spatial prediction (Lehmann et al. 2003), multivariate adaptive regression splines (Friedman 1991), Bayesian approach (Tucker et al. 1997), artificial neural networks (Pearson et al. 2002; Thuiller 2003; Thuiller et al. 2009), climatic envelopes (Busby 1991; Carpenter et al. 1993; Hijmans et al. 2001), generalised dissimilarity modelling (Ferrier et al. 2007), genetic algorithms (Stockwell & Peters 1999), classification and regression trees (Thuiller 2003; Thuiller et al. 2009), ecological niche factor analysis (Hirzel et al. 2002; David & Stockwell 2006), maximum entropy (Phillips et al. 2006), and support vector machines (Vapnik & Izmailov 2018).

Among these approaches, some employed the use of abundance (presence only, or presence and absence/background) data for the prediction of species' distribution. The background data are pseudo-absence points generated using the random point tool in a Geographic Information System (GIS) software environment. Different forms of the abundance data have been adopted in various studies on habitat suitability modelling of waterbirds. For example, maximum entropy (Vallecillo et al. 2016) was applied to investigate the factors influencing the distribution of waterbirds. Elith et al. (2006) and Aguirre-Gutiérrez et al. (2013) argued that the maximum entropy method offered better predictive performance, most especially for presence-only data using background data as well for model performance evaluation. This method uses background data to account for sampling bias and resolve the problem of non-detectability/false absences/false negatives associated with mobile organisms (Phillips et al. 2009) such as waterbirds. Based on this premise, this study explored the integration presence and absence data for waterbirds with associated environmental variables to predict their habitat suitability using the Automatic Linear Modelling Algorithm (ALMA) and GIS methods.

The Automatic Linear Modelling Algorithm, introduced by Yang (2013), is a form of a regression model with various model selection methods (such as forward stepwise, best subsets, include all predictors) to rank the independent variables based on the computation of a predictor importance coefficient (referred to as “importance ratios” in this study). These predictor importance coefficients are relative values, that sum up to 1.0 and each is ranked as an independent (predictor) variable according to its importance in the model (IBM Corporation 2011). ALMA is an extension of the multiple linear regression functions in the Statistical Package for Social Sciences (SPSS) software (IBM Corporation 2011) that can explore both the species' presence and absence data, but which has not previously been applied in any studies related to habitat suitability modelling. In this study, the model selection method (include all predictors) was selected to generate the predictor importance coefficients of all the environmental variables used. The novel variable selection capability of ALMA was integrated with geospatial techniques to determine the influence of eco-climatic variables on the occurrence of the waterbirds. The research aimed at comparing the population estimate, and developing robust habitat suitability models towards improved conservation of waterbirds in the wetlands (PIW and PW) of Peninsular Malaysia.

MATERIALS AND METHODS

1) Study areas

The study was undertaken in wetlands at Paya Indah (PIW) and Putrajaya (PW), Peninsular Malaysia (Figure 1). PIW is a natural wetland (101°36′23″E to 101°36′51″E and 2°51′21″N to 2°51′35″N), adjacent to the administrative area of Putrajaya (Rajpar et al. 2017). It covers an areaof 3,050 ha, of which 450 ha are managed by the Department of Wildlife and National Parks, Peninsular Malaysia (Lafferty et al. 2013). PIW comprises 14 lakes, with both disturbed and undisturbed peat swamp forest (Leveau et al. 2018), and five predominant land cover classes: marshy swamps, a lotus swamp, a lake, an open area with scattered trees, and scrub (Rajpar et al. 2017). Among the 24 species of waterbirds recorded in the wetland, the Purple Swamphen Porphyrio porphyrio indicus is the dominant species (Zakaria & Rajpar 2010; Rajpar & Zakaria 2011).

Fig. 1.

Paya Indah and Putrajaya wetlands in Peninsular Malaysia. LULC=Land cover, NDVI=Normalised Difference Vegetation Index, NDWI=Normalised Difference Water Index, ALMA=Automatic Linear Modelling Algorithm.

img-z3-8_39.jpg

PW is an artificial wetland (101°41′54″E to 101°42′26″E and 2°57′43″N and 2°57′49″N) in Putrajaya, Peninsular Malaysia (Rajpar & Zakaria 2013). It covers an area of 200 ha with five land cover classes: planted areas, open water, islands, inundated areas, and roads. PW wetland comprises 24 lakes that primarily control the water level and trap pollutants derived from upstream sources flowing into the catchment areas of the Chua and Bisa rivers. PW supports four vegetation classes: aquatic plants including emergent plants, fruiting trees, flowering trees and bushes, and shrubs (Rajpar & Zakaria 2013).

The construction of the habitat suitability models for these wetlands involved five stages, detailed below (see also Figure 2):

Fig. 2.

Framework for habitat suitability modelling of Waterbirds in Paya Indah wetland and Putrajaya Wetlands of Peninsular Malaysia.

img-z4-1_39.jpg

2) Bird surveys and background data of waterbirds

To obtain the occurrence data of all waterbird species in Peninsular Malaysia, we used the distance sampling point count technique. This technique was employed to survey birds at PW and PIW from November 2016 to January 2019. This technique is less demanding and more proficient than other sampling techniques for determining the status of avian populations (Rajpar & Zakaria 2010; Martins et al. 2017). It involves identification of birds by sight and sound with settled or variable radius plots, and gives critical data on species abundance, diversity and density among various natural surroundings (Koli 2014; Ma & Wang 2015; Manjari & Withanage 2015). This method enhances deductibility, which permits the estimation of density and abundance of birds (Nadeau et al. 2008). A total of 57 count stations were established at PIW and 54 at PW; they were located at least 100 m apart to avoid double counting and chosen based on visibility of the surroundings using binoculars. Bird count surveys at each count station, with a maximum variable radius of 100 m, were conducted for 10 minutes from 0730–1100 am. Surveys were carried out four times each week over 26 consecutive months. Bird surveys were conducted 20 times at each point count station, and combined for analysis. The distance between the observer and the waterbirds observed from each count station was measured using a Hypsometer (TruePulse R 200x model). The method was followed as described by Nadeau et al. (2008), Rajpar and Zakaria (2010), Martins et al. (2017), and Mohamed and Anjana (2017). The method efficiently ensures bias reduction with improved data accuracy and precision. Hutto and Young (2002) recommended ten-minute counts to reduce the numbers of birds ignored.

The following data were recorded during the survey: lake, species observed on the lake, the total number of individuals observed, vegetation type, land use and time sighted. A set of 1,000 background points were generated using the “create random points” tool of ArcGIS 10.4 to account for sampling bias (Phillips et al. 2006; Phillips et al.2009).

3) Estimation of waterbirds species diversity, richness and evenness

The waterbirds species diversity, richness and evenness indices (Equations 13) were computed and developed using EcoIndR Version 1.4 package (Gonzalez 2018) in R Software Version 3.5.2 (R Core Team 2019). The independent t-test was used to determine if significant differences (p<0.05) existed in the diversity, richness and evenness of the waterbirds species between PIW and PW. The details of the waterbirds species diversity, richness and evenness indices used are as follows:

e01_39.gif

Where, S=Shannon's diversity index, n=Number of species, i=Abundance of species,

N=Total number of all individuals and pi is th relative abundance of each species

e02_39.gif

Where, N=Sum of species in the plot, n=Sum of individuals of all species, M=Margalef's richness index

e03_39.gif

Where, E=Pielou's J evenness index, S=Shannon's diversity index, n=Number of species

4) Acquisition of satellite images, climatic/hydrology data and collection of ground truth points

Satellite images and climatic/hydrology data were acquired to provide the primary data required for the explanatory variables for the modelling process. Sentinel 2A MSI Level-1C satellite images were sourced from United States Geological Survey (USGS) archives via Global Visualisation Viewer at scales of 10 m resolutions. These images were captured during the driest month (April 2018) to minimise the interference from cloud cover and depict the real state of land cover of the sites. The acquired satellite data was used as the principal source for the extraction of land cover (LULC), Normalised Difference Vegetation Index (NDVI) and Normalised Difference Water Index (NDWI). Climatic data (relative humidity, rainfall, wind speed, atmospheric pressure and atmospheric temperature) were obtained from the National Climate Center, Malaysian Meteorological Department, Malaysia. Hydrological data (electrical conductivity, dissolved oxygen, water quality index (Equation 4) (House & Ellis 1987; Breaban et al. 2012) were obtained from the Department of Wildlife and National Parks, Peninsular Malaysia. Data on turbidity, temperature, salinity, pH, minimum depth and maximum depth were obtained from the National Hydraulic Research Institute of Malaysia.

Ground truth points were required for model validation. Field surveys were carried out for the purposes of accurate assessment of land cover classes (ground truthing). A hand-held GPS (GPS Map 78s, GARMIN) was used to collect coordinates of points representing the different land cover classes. Ground truthing, at 190 ground control points, was performed in the study area during April 2018, during the season similar to the acquisition of satellite datasets. The ground control points were used as training samples during the land cover classification of the wetlands. Error matrices and kappa statistics were computed using the accuracy assessment tool in ERDAS Imagine (2014) software.

e04_39.gif

Where, Qi=sub-index for ith water quality parameter, Wi=weight associated with ith water quality parameter, n=number of water quality parameters.

5) Creation of factor files and maps

We used four criteria (hydrology, climate, waterscape, landscape) and 17 factors (water temperature, pH, Dissolve oxygen, electrical conductivity, salinity, turbidity, maximum depth, minimum depths, water quality index, atmospheric pressure, wind speed, rainfall, relative humidity, atmospheric temperature, Normalised Difference Water Index, Normalised Difference Vegetation Index and landcover) to model the influence of eco-climatic factors on the spatial heterogeneity of waterbirds in Peninsular Malaysia. The selection of two factors for landscape (land cover and Normalised Difference Vegetation Index [a measure of vegetation cover, forage availability and human activity]) and one for waterscape (Normalised Difference Water Index [a measure of water availability]) as suitability factors was based on Hirzel et al. (2001); Brotons et al. (2004); Tian et al. (2008); Khoury et al. (2010); Tang et al. (2011). The Normalised Difference Vegetation and Normalised Difference Water indices were extracted from the Sentinel 2A imagery. Pixel-based image classification, using the supervised classification method, was employed to determine the land cover (a measure of safe shelter and forage availability for waterbirds) of the wetlands. The United States Geological Survey (USGS) Land Classification Scheme (Anderson et al. 1976) was used as reference, although for our analysis we reduced the nine land cover classes to five at PIW and four at PW based on current land cover scenarios there. The five classes at PIW were semi-closed secondary forest, shrubland, marsh grassland (grasses, marsh and Lotus swamp), lakes, and urban areas (bare ground and built-up areas), and the four at PW were partly closed secondary forest, aquatic herbaceous vegetation and marsh swamp or aquatic grassy vegetation, lakes, and urban areas. The factor files (continuous and raster data) of the hydrological and climatic parameters were created using the inverse difference interpolation method according to the procedure of Knight et al. (2005).

6) Weighting and standardisation of factor files

We extracted the associated float values for the 17 continuous factor raster maps using the “extract to points” tool in ArcGIS 10.4 software and abundance data for waterbirds. Then we employed the Automatic Linear Modelling Algorithm (ALMA; Yang 2013) using the “include all predictors” model selection method to determine the predictor importance (importance ratio) for each factor using SPSS software (Version 20) (IBM Corporation 2011). The presence or absence data (number of individual waterbirds detected or not detected, 75%) served as the dependent variable (see Equation 5). The importance ratio of each factor was standardised into a weight value (in percentage) using Equation 6 and then used to rank the factors indicating their level of influence.

e05_39.gif

Where,

yi=Waterbird Abundance

βo=Constant term

β1 to β17=Coefficients relating to the eco-climatic factors

ei=Error term with mean value of 0

x1=Electrical conductivity, x2=Dissolved oxygen, x3=Water Quality Index, x4=Turbidity, x5=Water temperature, x6=Salinity, x7=pH, x8=Minimum depth, x9=Maximum depth, x10=Relative humidity, x11=Rainfall, x12=Wind speed, x13=Atmospheric pressure, x14=Atmospheric temperature, x15=NDWI, x16=NDVI, x17=LULC

e06_39.gif

7) Habitat suitability model development and validation

The 17 factors and suitability map were categorized into five suitability classes (highly suitable, suitable, moderately suitable, non-suitable and highly non-suitable) to determine the habitat suitable for the waterbirds. The categorization based on the habitat suitability continuum framework developed by Dong et al. (2013) using the Jenk's natural breaks (Jenk 1967) via the Reclassify tool of ArcGIS 10.4. These reclassified factor layers were overlaid based on their weight values (%) to create a habitat suitability map for waterbirds. Twenty-five per cent of the abundance data of the waterbirds and the 1,000 pseudo-absence points were employed to validate the models (Philipps et al. 2006). The “extract to points” tool in ArcGIS 10.4 software was used to extract the associated suitability code (0 or 1) from the re-classified suitability maps to these presence or absence and background data.

Validated points and predicted map attributes from ArcGIS were exported to Microsoft Excel 2017 (in a “text tab-delimited layer format”), then imported into the R 3.5.2 software environment for receiver operating characteristic analysis (ROC). ROC analysis was employed to compute the area under the curve (AUC) and evaluate model performance using the pROC package (Robin et al. 2011). The area under the curve, from the plot of sensitivity (true positives) against 1 minus specificity, depicted the model's capacity to discriminate presence from absence. According to Hosmer andLemeshow (2000), AUC valuesof 0.50–0.70, 0.70–0.90, and greater than 0.9 signify that a model's accuracy is low, moderate or high. Even, AUC values >0.75 are considered accurate, acceptable and suitable for predicting species distribution (Elith 2000; Vanagas 2004; Phillips et al. 2006; Lobo et al. 2008).

RESULTS

We recorded 31 waterbird species at the PIW and PW wetlands (see Table 1). We observed more waterbirds in PW (n=28,303), with greater species diversity (S=7.60) and greater richness (M=267.30), than in PIW (n=19,160; S=7.10; M=156.00). The species with the highest abundance at PW were [Black-crowned night heron] Nycticorax nycticorax, [Grey heron] Ardea cinerea, [Great egret] Casmerodius albus, [Lesser whistling duck] Dendrocygna javanica, [Black bittern] Dupetor flavicollis and [Cinnamon bittern] Ixobrychus cinnamomeus. PIW also had some common species such as [Black-winged kite] Elanus caeruleus, [Green iora] Aegithina virdissima, [Pied Triller] Lalage nigra, and [Large-billed crow]

Table 1.

Waterbirds and their relative abundances in the Paya Indah (PIW) and Putrajaya (PW) wetlands

img-z7-2_39.gif

Fig. 3.

The performance of the fitted habitat suitability models for waterbirds in Paya Indah and Putrajaya wetlands of Peninsular Malaysia.

img-z8-1_39.jpg

Corvus macrorhynchos.

The performance of the fitted habitat suitability models revealed that both models performed robustly (Figure 3) with AUC values significantly greater than 0.50 of a random model. The AUC value for waterbirds in PIW was 0.960 and in PW 0.961. Habitat suitability evaluation of the PIW and PW wetlands was based on automatic linear modelling (see Table 2).

Results from PIW, show that salinity made the greatest contribution to the habitat suitability model for waterbirds with an importance ratio of 0.317 and a weight of 31.08%. However, the contributions of eleven eco-climatic factors (Electric conductivity, Dissolved Oxygen, Turbidity, Water temperature, Salinity, Minimum Depth, Maximum Depth, Relative humidity, Rainfall, Wind speed and Atmospheric temperature) to the habitat suitability model for PIW were highly significant (p ≤ 0.05) based on automatic linear modelling. Specifically, the ranges of Electric conductivity (15.49 uS/cm to 41.18 uS/cm), Dissolved Oxygen (4.47–8.22 mg/L), turbidity (2.02–23.73 NTU), Water temperature (24.45–30.79 °C), Salinity (0.50–5.04 ppt), minimum water depth (0.65–5.92 m), Maximum depth (3.12–20.74m), Relative humidity (27.855–77.530%), Rainfall (9.976– 10.691 mm), Wind speed (1.487–1.618 m/s) and Atmospheric temperature (27.741–27.773 °C) offered suitable habitats for this species (Table 3).

Results from PW show that land cover (a measure of vegetation cover, forage availability and human activity) contributed most to the habitat suitability model of waterbirds there, with an importance ratio of 0.334 and a weight 33.4%. Ten eco-climatic factors (Electric conductivity, Dissolved Oxygen, Turbidity, Water temperature, Salinity, pH, Rainfall, Atmospheric pressure, Atmospheric temperature and LULC) contributed significantly (p ≤ 0.05) to the habitat suitability model of waterbirds in PW. In particular, the range of Electric conductivity (1009.436–1009.935 Hpa), Dissolved Oxygen (6.12–7.35mg/L), Turbidity (12.67–76.85 NTU), Water temperature (29.94–30.72 °C), Salinity (0.03–0.08 ppt), pH (7.35–7.58), Rainfall (8.53–9.03 mm), Atmospheric pressure (1009.44–1009.94 Hpa), Atmospheric temperature (27.309–27.564 °C) and LULC offered suitable habitats for waterbirds in PW (Table 3). Habitat suitability models were generated for waterbirds in PIW and PW (Figure 4 and Table 4). The habitat suitability map for waterbirds at PIW, showed that the area considered moderately suitable occupied the greatest area (535.40 ha; 33.90%), while the area deemed highly suitable occupied the smallest area (103.55 ha; 6.56%). The habitat suitability map for waterbirds at PW showed that the area considered suitable occupied the greatest area (430.28 ha; 30.29%), whereas the area deemed highly non-suitable for waterbirds occupied the smallest area (0.84 ha; <1%).

DISCUSSION

Our study revealed that most climatic and hydrological factors significantly influenced the occurrence and distribution of waterbirds in PIW and PW, Peninsular Malaysia. In contrast, waterscape (Normalised Difference Water Index) and landscape (Normalised Difference Vegetation Index) did not significantly influence waterbird distribution in either wetland. The important contribution of the climatic (relative humidity, rainfall, wind speed, atmospheric temperature) and hydrological factors (electric conductivity, dissolved oxygen, turbidity, pH, water temperature, salinity, minimum depth, maximum depth) to the abundance of waterbirds in PIW confirmed Rajpar and Zakaria (2014) and Wormworth and Mallon (2014) who previously showed that some of these factors were very important for waterbirds in choosing their habitat, while considering that there was a negative relationship between waterbirds and atmospheric temperature and relative humidity at PIW. Wormworth and Mallon (2014) had previously shown that rainfall and wind are limiting factors affecting the distribution of waterbirds due to their direct impact on wetland habitats.

Table 2.

Factors used for habitat suitability evaluation by the Automatic Linear Modelling Algorithm and their weights of importance for waterbirds in Paya Indah (PIW) and Putrajaya (PW) wetlands

img-z9-2_39.gif

Land cover at PW contributed considerably to the habitat suitability for waterbirds (and much more so than at PIW) showing the role that anthropogenic activities (the main driver of land cover), forage availability, and vegetation cover play in waterbird occurrence and distribution. Van Niekerk (2010) and Mundava et al. (2012) considered that anthropogenic pressure poses a serious threat to the population growth of waterbirds. PIW and PW experience various forms of human activities because they are situated within highly urbanized areas (the former in Selangor State and the latter in Putrajaya Federal territory) of Peninsular Malaysia. Zakaria and Rajpar (2013) and Hassen-Aboushiba (2015) reported tin mining, agricultural activities, and tourism infrastructural development to be the major anthropogenic activities affecting PIW. Urban sprawl and water purification and supply may have contributed to the landscape dynamics and variation affecting Putrajaya Wetland.

This study supported the work of Rajpar and Zakaria (2014), in Putrajaya Wetland, Malaysia, and Jahanbakhsh et al. (2017), in Parishan International wetland, Iran, in finding that vegetation cover affected the habitat selection, distribution and diversity of waterbirds. The influence of climate change on wetland vegetation composition, structure, hydro-morphology, and consequently the distribution and sustainability of waterbird populations in wetlands cannot be underestimated. According to Porte and Gupta (2017), this global phenomenon has broad impacts on the distribution, morphology, carrying capacity and seasonal variation of wetlands which in turn affect the feeding and breeding activities of waterbirds.

Table 3.

Environmental attributes of the Paya Indah (PIW) and Putrajaya (PW) wetlands

img-z10-2_39.gif

In our study, the greater suitability of the habitat of Putrajaya wetlands for the occurrence and distribution of waterbirds can be attributed to the significant importance of water pH, atmospheric pressure and land cover. In contrast, we revealed that the influence of relative humidity and wind speed on the habitat of PIW was negligible. Atmospheric pressure had been shown to be a contributory factor affecting the distribution of several bird species; for example, Metcalfe et al. (2013) showed that atmospheric pressure predicted accurately the distribution of the White-throated Sparrow Zonotrichia albicollis in Canada. We consider that there may be a link between the local atmospheric temperatures at ground level (a characteristic of the heat island effect of urban sprawl) to evaporation (Fukatani et al. 2016; Mirzaei et al. 2020). Mahdi et al. (2020) opined that a higher vegetation cover reduced the urban heat effects. The decreased atmospheric temperature that increased the atmospheric pressure reduces surface evaporation (Fukatani et al. 2016; Ozgur & Kacak, 2015). For this reason, water retention in the lakes at PW is higher than PIW ensuring water availability for forage growth. Therefore, the higher vegetation cover of PW than PIW contributed to the significant influence of atmospheric pressure on the waterbird species.

Fig. 4.

Habitat suitability models of waterbirds in Paya Indah and Putrajaya wetlands of Peninsular Malaysia.

img-z11-1_39.jpg

Table 4.

Attributes of habitat suitability models for waterbirds in Paya Indah (PIW) and Putrajaya (PW) wetlands

img-AqcMq_39.gif

Our findings disagreed with Russell et al. (2014) who found that turbidity had no significant influence on the distribution of waterbirds in the Wilderness Lake Complex of South Africa. At PIW, most lakes are less turbid than PW, open and deep. The low turbidity may be attributed to the dense aquatic vegetation at PIW, as Austin et al. (2017) have asserted that aquatic vegetation may minimize water turbidity through water flow reduction, stabilisation of sediments and nutrient competition with phytoplankton. Our study, rather like that of Rajpar and Zakaria (2011), found that water depth influenced the occurrence, diversity and distribution of waterbirds in PIW. Fluctuating water levels may affect habitat characteristics that in turn cause sudden changes in waterbird communities (Johnson et al. 2007). In addition to limiting access to foraging habitats, water depth affects the net energy intake of waterbirds, especially wading birds, because their foraging efficiency is reduced by increasing water depth. Gawlik (2002) indicated that the locomotion of wading birds foraging for prey in the water column may be slowed in deep water because water resistance increases with depth.

Generally, PW is characterized by slow-flowing waters that contribute to its suitability for waterbirds (Rajpar & Zakaria 2014). Several of the lakes in the wetlands are shallow due to siltation with slow-flowing waters. For this reason, waterbirds were mostly and currently distributed in shallow lakes. Furthermore, water temperature is a controlling factor that affects the functioning of the aquatic ecosystem, and influences the growth and distribution of flora and fauna on which most waterbirds depend for their nourishment (Dwivedi & Pandey 2002; Singh & Mathur 2005; Jalal & Sanalkumar 2012; Tank & Chippa 2013). Water temperature determined the distribution pattern of macrophytes, wetland productivity and waterbirds composition, which vary with depth, season and geographical location. At PIW and PW, it was noted that lower water temperature significantly contributed to the higher abundance of waterbirds. This observation supported the views of Deshkar et al. (2010) that low water temperature is more suitable to waterbirds than high water temperature.

The pH of water bodies affects productivity and habitat diversity (Miller et al. 1986), with alkaline waters supporting greater macro-invertebrate populations than acid waters and thereby attracting more waterbirds to alkaline water bodies (Longcore et al. 2006). At both PIW and PW, the water was slightly alkaline (7.2–9.2 pH) during the entire study period, although pH did not contribute significantly to waterbird population size in PIW. Longcore et al. (2006) established a significant influence of pH with species richness of phytoplankton, and the waterbirds, which depend on these organisms. Furthermore, electrical conductivity (a measure of the ionic composition of water) plays an essential role in water productivity and growth of aquatic plants consumed by some waterbirds (Manikannan et al. 2012). This water parameter significantly influenced the distribution of waterbirds in both PIW and PW. Our findings aligned with the opinion of Pandiyan (2002) who considered that electrical conductivity significantly affected the population structure of waders (Charadriiformes) in Tamil Nadu, India.

Moreover, the significance of dissolved oxygen for waterbird distribution cannot be underestimated in either PIW or PW wetlands. According to Sathe et al. (2001), dissolved oxygen is vital in the regulation of the metabolic processes of aquatic plants and an indicator of aquatic ecosystem health. In previous studies, a significant relationship was found to exist between dissolved oxygen and waterbird populations (Thapa & Saund 2012; Sulai et al. 2015; Haq et al. 2018). Therefore, hydrology parameters are considered among the most important factors capable of influencing the aquatic environment and in turn determining the abundance of waterbirds in wetlands, as shown in our study. In general, based on our habitat suitability models, PW offered a larger area of habitat suitable for the occurrence, distribution, survival and perpetuity of waterbird populations, than did PIW. Despite Putrajaya wetland's proximity to human habitation, its favourable water pH, atmospheric pressure and land cover led to its considerable potential as waterbird habitat.

CONCLUSION

Of the two wetland areas, PW provided more suitable habitat than PIW for waterbird populations. The difference between them can be attributed to the favourable water pH, atmospheric pressure and land cover (providing suitable vegetation for foraging) of PW. However, landscape, waterscape, climatic and hydrological factors significantly influenced the occurrence and distribution of the species occurring at PIW and PW as revealed by the models. In particular, the land cover influenced the distribution of waterbird species in Putrajaya Wetlands. Thus, the higher the climatic (atmospheric pressure, relative humidity) and hydrological (electrical conductivity, turbidity) variables, the waterbirds populations of the wetlands were more diverse, more widely distributed and more greatly occupied. The habitat suitability models we have developed helped clarify the spatial heterogeneity of waterbird populations in relation to patterns of eco-microclimatic factors and habitat use. It is believed that the adoption of the models as management tools can help in the maintenance of the wetlands' habitat quality and support effective management of waterbirds species.

ACKNOWLEDGMENTS

We would like to thank the Department of Wildlife and National Parks, Peninsular Malaysia for permission to conduct this study. This research was partially funded by the Putra Grant (GP-IPS/2018/9638000), Universiti Putra Malaysia, Selangor, Malaysia.

REFERENCES

1.

Aguirre-Gutiérrez J, Carvalheiro LG, Polce C, van Loon EE, Raes N, Reemer M & Biesmeijer JC (2013) Fit-for-Purpose: Species Distribution Model Performance Depends on Evaluation Criteria – Dutch Hoverflies as a Case Study. PLoS ONE 8: e63708.  https://doi.org/10.1371/journal.pone.0063708  Google Scholar

2.

Ajonina G, Amougou J, Ayissi I, Ajonina P, Ntabe E & Dongmo M (2009) Waterbirds as bio-indicators of seasonal-climatic changes in river basin properties: eight years monthly monitoring in lower Sanaga, Cameroon. IOP Conference Series: Earth and Environ Sci 6: 292021. Google Scholar

3.

Al-Saadi LM, Jaber SH & Al-Jiboori MH (2020) Variation of urban vegetation cover and its impact on minimum and maximum heat islands. Urban Climate 34: 100707 Google Scholar

4.

Amat JA & Green AJ (2010) Waterbirds as bio indicators of environmental conditions. In: Hurford C, Schneider M, Cowx I (eds) Conservation monitoring in freshwater habitats. pp 45–52. Springer, Dordrecht. Google Scholar

5.

Anderson JR, Hardy EE, Roach JT & Witmer RE (1976) A Land Use and Land Cover Classification System for Use with Remote Sensor Data. Geological Survey, Washington DC. Available at www.scirp.org(accessed on February 24, 2020). Google Scholar

6.

Austin ÅN, Hansen JP, Donadi S & Eklöf JS (2017) Relationships between aquatic vegetation and water turbidity: A field survey across seasons and spatial scales. PLoS ONE 12: e0181419.  https://doi.org/10.1371/journal.pone.0181419  Google Scholar

7.

Breaban IG, Gheţeu D & Paiu M (2012) Determination of Water Quality Index of Jijia and Miletin Ponds. Bull UASVM Agric 69: 1. Google Scholar

8.

Brotons L, Thuiller W, Miguel B & Hirzel AH (2004) Presence-absence versus presence-only modelling methods for predicting bird habitat suitability. Ecography 27: 437–448. Google Scholar

9.

Burger J & Eichhorst B (2005) Heavy metals and selenium in grebe eggs from Agassiz National Wildlife Refuge in northern Minnesota. Environ Monit Assess 107: 285–95. Google Scholar

10.

Busby JR (1991) Bioclim, a Bioclimatic Analysis and Prediction System. In: Margules CR and Austin MP (eds) Nature Conservation: Cost Effective Biological Surveys and Data Analysis. pp 64–68. CSIRO, Canberra. Google Scholar

11.

Carpenter G, Gillison AN & Winter J (1993) A flexible modelling procedure for mapping potential distributions of plants and animals. BiodiversConserv2: 667–680. Google Scholar

12.

David RB & Stockwell D (2006) Improving ecological niche models by data mining large environmental datasets for surrogate models. Ecol Model 192: 188–196. Google Scholar

13.

Deshkar S, Rathod J & Padate G (2010) Avifaunal diversity and water quality analysis of an inland wetland. J Wet Eco 4: 1–32. Google Scholar

14.

Dong Z, Wang Z, Liu D, Li L, Ren C, Tanga X et al. (2013) Assessment of habitat suitability for waterbirds in the West Songnen Plain, China, using remote sensing & GIS. EcolEng 55: 94–100. Google Scholar

15.

Dwivedi BK & Pandey GC (2002) Physico-chemical factors and algal diversity of two ponds, (GirijaKund and Maqubara pond), Faizabad. Pollut Res 21: 361–370. Google Scholar

16.

Elith J (2000) Quantitative methods for modelling species habitat: comparative performance and an application to Australian plants. In: Ferson S & Burgman M (eds) Quantitative methods for conservation biology. pp 39–58. Springer, New York. Google Scholar

17.

Elith J, Graham CH, Anderson RP, Dudík M, Ferrier S, Guisan A et al. (2006) Novel methods improve prediction of species' distributions from occurrence data. Echography 29: 129–151. Google Scholar

18.

ERDAS (2014) ERDAS Imagine 2014. Hexagon Geospatial, Peachtree Corners Circle Nor-cross. Scientific Research Publishing. (n.d.).  www.Scirp.Org . Available at https://www.scirp.org/(S(351jmbntvnsjt1aadkposzje))/reference/ReferencesPapers.aspx?ReferenceID=1557945(accessed 23 December 2019). Google Scholar

19.

Fasola M, Rubolini D, Merli E, Boncompagni E & Bressan U (2009) Long-term trends of heron and egret populations in Italy, and the effects of climate, human-induced mortality, and habitat on population dynamics. Popul. Ecol. 52: 59–72. Google Scholar

20.

Ferrier S, Manion G, Elith J, & Richardson K (2007) Using generalized dissimilarity modelling to analyse and predict patterns of beta diversity in regional biodiversity assessment. Divers Distrib 13: 252–264. Google Scholar

21.

Franklin J (1995) Predictive vegetation mapping: geographic modelling of biospatial patterns in relation to environmental gradients. Prog Phys Geog 19: 474–499. Google Scholar

22.

Friedman JH (1991) Multivariate Adaptive Regression Splines. Ann Stat 19: 1–141. Google Scholar

23.

Fukatani Y, Orejon D, Kita Y, Takata Y, Kim J & Sefiane K (2016) Effect of ambient temperature and relative humidity on interfacial temperature during early stages of drop evaporation. Physical Review E 93(4): 04310.  https://doi.org/10.1103/PhysRevE.93.043103  Google Scholar

24.

Gawlik DE (2002) The effects of prey availability on the numerical response of wading birds. Kushlan, JA 2000a. Heron feeding habitat conservation. In: Kushlan JA & Hafner H (eds) Heron Conservation. pp 219–235. Academic Press, San Diego. Google Scholar

25.

Gonzalez CG (2018) EcoIndR: Ecological Indicators Package. R package version 1.4. Available at https://CRAN.R-project.org/package=EcoIndR . (accessed on 21 December 2018) Google Scholar

26.

Guisan A & Zimmermann NE (2000) Predictive habitat distribution models in ecology. Ecol Model 135: 147–186. Google Scholar

27.

Gumbricht T, Roman-Cuesta RM, Verchot LV, Herold M, Wittmann F, Householder E et al. (2017) Tropical and Subtropical Wetlands Distribution version 2.  https://doi.org/10.17528/cifor/data.00058  Google Scholar

28.

Haq RU, Eiam-Ampai K, Ngoprasert D, Sasaki N & Shrestha RP (2018) Changing Landscapes and Declining Populations of Resident Waterbirds: A 12-Year Study in Bung Boraphet Wetland, Thailand. Trop Conserv Sci 11: 1–17. Google Scholar

29.

Hassen-Aboushiba, AB (2015) Assessing the Effects of Aquatic Vegetation Composition on Waterbirds Distribution and Richness in Natural Freshwater Lake of Malaysia. Amer J Life Sci3: 316–321. Google Scholar

30.

Hijmans RJ, Guarino L, Cruz M & Rojas E (2001) Computer tools for spatial analysis of plant genetic resources data 1. DIVA-GIS. Plant Genet Resour News 127: 15–19. Google Scholar

31.

Hirzel AH, Hausser J, Chessel D & Perrin N (2002) Ecological niche factor analysis: How to compute habitat suitability maps without absence data. Ecology 83: 2027–2036. Google Scholar

32.

Hirzel AH, Helfer V & Metral F (2001) Assessing habitat-suitability models with a virtual species. Ecol Model 145: 111–121. Google Scholar

33.

Hirzel AZ & Lay GL (2008) Habitat suitability modelling and niche theory. J Appl Ecol45: 1372–1381. Google Scholar

34.

Ho CS (2006) Putrajaya Administrative Centre of Malaysia–Planning Concept and Implementation. Sustainable Urban Development and Governance Conference at Sung Kyun Kwan University Seoul on 16 Nov 2006. Pp20. Google Scholar

35.

Hosmer D & Lemeshow S (2000) Applied Logistic Regression. p 392. Wiley, New York. Google Scholar

36.

House MA & Ellis JB (1987) The development of water quality indices for operational management. Water Sci Technol 19: 145–154. Google Scholar

37.

Hutto RL & Young JS (2002) Regional landbird monitoring: perspectives from the Northern Rocky Mountains. Wildl Soc Bull 30: 738–750. Google Scholar

38.

IBM SPSS Statistics 26 (2019) Available at https://www.ibm.com/support/pages/downloading-ibm-spssstatistics-26(accessed on 6 December 2019). Google Scholar

39.

Ismail A & Rahman F (2013) Does weather play an important role in the early nesting activity of colonial waterbirds? A case study in Putrajaya Wetlands, Malaysia. Trop Life Sci Res 24: 1–7. Google Scholar

40.

Jahanbakhsh GM, Khorasani N, Morshedi J, Danehkar A & Naderi M (2017) Factors influencing the abundance and species richness of overwintered waterbirds in Parishan International Wetland in Iran. Appl Ecol Env Res15: 1565–1579. Google Scholar

41.

Jalal FN, Sanalkumar MG (2012) Hydrology and water quality assessment of Achencovil river in relation to pilgrimage season. Int J Sci Res 2: 1–5. Google Scholar

42.

Jenks GF (1967) The data model concept in statistical mapping. International Yearbook of Cartography 7: 186–190. Google Scholar

43.

Johnson KG, Allen MS & Havens KE (2007) A review of littoral vegetation, fisheries, and wildlife responses to hydrologic variation at Lake Okeechobee. Wetlands 27: 110–126. Google Scholar

44.

Khoury M, Higgins J & Weitzell R (2010) A freshwater conservation assessment of the Upper Mississippi River basin using a coarse- and fine-filter approach. Freshwater Biol 56: 162–179. Google Scholar

45.

Knight Y, Yu B, Jenkins G & Morris K (2005) Comparing rainfall interpolation techniques for small subtropical urban catchments. Modelling and Simulation Society of Australia and New Zealand, Conference Publication, December, 2005. Google Scholar

46.

Koli VK (2014) Diversity and status of avifauna in Todgarh-Raoli Wildlife Sanctuary, Rajasthan, India. J Asia Pac Biodivers 7: 401–407. Google Scholar

47.

Lafferty KD, Rodriguez DA & Chapman A (2013) Temporal and spatial variation in bird and human use of beaches in southern California. Springer Plus 2: 38. Google Scholar

48.

Lehmann A, Overton JM & Leathwick JR (2003) GRASP: generalized regression analysis and spatial prediction. Ecol Model 160: 165–183. Google Scholar

49.

Leveau LM, Isla FI & Bellocq MI (2018) Predicting the seasonal dynamics of bird communities along an urban-rural gradient using NDVI. Landsc Urban Plan 177: 103–113. Google Scholar

50.

Lobo JM, Enez-Valverde AJ & Real R (2008) AUC: a misleading measure of the performance of predictive distribution models. Glob Ecol Biogeogr 17: 145–151. Google Scholar

51.

Longcore JR, McAuley DG, Pendelton GW, Bennatti CR, Mingo TM, & Stromborg KL (2006) Macroin-vertebrate abundance, water chemistry, and wetland characteristics affect use of wetlands by avian species in Maine. Hydrobiologia 567: 143–167. Google Scholar

52.

Ludwig GX, Alatalo RV, Helle P, Linden H, Lindstom J & Siitari H (2006) Short and long term population dynamical consequences of asymmetric climate change in Black Grouse. P Roy Soc B-Biol Sci 273: 2009–2016. Google Scholar

53.

Ma G & Wang S (2015) Temporal and spatial distribution changing characteristics of exogenous pollution load into Dianchi Lake, Southwest of China. Environ Earth Sci 74: 3781–3793. Google Scholar

54.

Malaysian Nature Society Bird Conservation Council (2015) MNS Conservation Publication. No.14 A Checklist of the Birds of Malaysia . 2nd ed . Malaysian Nature Society, Kuala Lumpur. Google Scholar

55.

Manikannan R, Asokan S, Mohamed A & Ali S (2012) Abundance and Factors Affecting Population Characteristics of Waders (Charadriiformes) in Great Vedaranyam Swamp of Point Calimere Wildlife Sanctuary, South-east Coast of India. Int J Ecosyst 2: 6–14. Google Scholar

56.

Manjari BJ & Withanage UC (2015) Variation of avifaunal diversity in relation to land-use modifications around a tropical estuary, the Negombo estuary in SriLanka. J Asia Pac Biodivers 8: 72–82. Google Scholar

57.

Martins CO, Rajpar MN, Nurhidayu S, & Zakaria M (2017) Habitat Selection of Dendrocygna javanica in Heterogeneous Lakes of Malaysia. J Biodivers Manage Forestry 6: 1–6. Google Scholar

58.

Matsinos YG & Wolf WF (2003) An individual-oriented model for ecological risk assessment of wading birds. Ecol Model 170: 471–8. Google Scholar

59.

Metcalfe J, Schmidt KL, Bezner-Kerr W, Guglielmo CG & MacDougall-Shackleton SA (2013) White-throated sparrows adjust behaviour in response to manipulations of barometric pressure and temperature. Anim-Behav 86: 1285–1290. Google Scholar

60.

Miller WW, Joung HM, Mahannah CN, Garrett JR (1986) Identification of water quality differences Nevada through index application. J Environ Qual 15: 265–272. Google Scholar

61.

Mirzaei M, Verrelst J, Arbabi M, Shaklabadi Z & Lotfizadeh M (2020) Urban Heat Island Monitoring and Impacts on Citizen's General Health Status in Isfahan Metropolis: A Remote Sensing and Field Survey Approach. Remote Sens 12: 1350.  https://doi.org/10.3390/rs12081350  Google Scholar

62.

Mohamed SK & Anjana P (2017) Conservation status, species composition, and distribution of Avian Community in Bhimbandh Wildlife Sanctuary, India. J Asia Pac Biodivers 10: 20–26. Google Scholar

63.

Morganti M, Manica M, Bogliani G, Gustin M, Luoni F, Trotti P et al. (2019) Multi-species habitat models highlight the key importance of flooded reedbeds for inland wetland birds: implications for management and conservation. Avian Res 10.  https://doi.org/10.1186/s40657-019-0154-9  Google Scholar

64.

Mundava J, Caron A, Gaidet N, Couto FM, Couto JT, De Garine-Wichatitsky M et al. (2012) Factors influencing long-term and seasonal waterbirds abundance and composition at two adjacent lakes in Zimbabwe. Ostrich 83: 69–77. Google Scholar

65.

Nadeau CP, Conway CJ, Smith BS & Lewis TE (2008) Maximizing detection probability of wetland-dependent bird during point count surveys in North-western Florida. Wilson J Ornithol 120: 513–518. Google Scholar

66.

Ozgur E & Koçak K (2015) The effects of the atmospheric pressure on evaporation. Acta Geo 1: 17–24. Google Scholar

67.

Pandiyan J (2002) Ecology of shorebirds in the tidal flats of Cauvery deltaic region of Southern India. Ph.D. Thesis, Bharathidasan University, Thiruchirappalli, India. Google Scholar

68.

Paolini KE, Strickland BK, Tegt, JL, VerCauteren KC & Street GM (2018) Seasonal variation in preference dictates space use in an invasive generalist. PLoS ONE13: e0199078.  https://doi.org/10.1371/journal.pone.0199078  Google Scholar

69.

Pearson RG, Dawson TP, Berry PM & Harrison PA (2002) SPECIES: A spatial evaluation of climate impact on the envelope of species. Ecol Model154: 289–300. Google Scholar

70.

Phillips SJ, Anderson RP & Schapire RE (2006) Maximum entropy modelling of species geographic distributions. Ecol Model190: 231–259. Google Scholar

71.

Phillips SJ, Dudik M, Elith J, Graham CH, Lehmann A, Leathwick J et al. (2009) Sample selection bias and presence-only models of species distributions. Ecol Appl 19: 181–197. Google Scholar

72.

Porte DS & Gupta S (2017) Assessment of distribution patterns of wetland birds between unpolluted and polluted ponds at Ratanpur, District Blaspur, Chhattisgarh, India. Indian J Sci Res 12: 204–215. Google Scholar

73.

R Core Team (2019) Available at R-project.org website: https://www.R-project.org/ .(accessed on 24 August 2019). Google Scholar

74.

Rahman F & Ismail A (2018) Waterbirds: An Important Bio-Indicator of Ecosystem. Pertanika J Sci Technol 4: 81–90. Google Scholar

75.

Rajpar MN & Zakaria M (2010) Density and Diversity of Water Birds and Terrestrial Birds at Paya Indah Wetland Reserve, Selangor Peninsular Malaysia. J Biol Sci 10: 658–666. Google Scholar

76.

Rajpar MN & Zakaria M (2011) Effects of Water Level Fluctuation on Waterbirds Distribution and Aquatic Vegetation Composition at Natural Wetland Reserve, Peninsular Malaysia. ISRN Ecol 2011: 1–13. Google Scholar

77.

Rajpar MN & Zakaria M (2013) Assessing an Artificial Wetland in Putrajaya, Malaysia, as an Alternate Habitat for Waterbirds. Waterbirds 36: 482–493. Google Scholar

78.

Rajpar MN, Zakaria M, Ozdemir I, Ozturk M & Gucel S (2017) Avian Assemblages at Paya Indah Natural Wetland Reserve, Malaysia. Expert Opin Environ Biol6: 1–6. Google Scholar

79.

Rajpar, MN & Zakaria M (2014) Effects of habitat characteristics on waterbirds distribution and richness in the wetland ecosystem of Malaysia. J Wildl and Parks28: 107–122. Google Scholar

80.

Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC et al. (2011) pROC: an open-source package for R and S+to analyze and compare ROC curves. BMC Bioinformatics, 12.  https://doi.org/10.1186/1471-2105-12-77  Google Scholar

81.

Rushton SP, Ormerod SJ & Kerby G (2004) New paradigms for modelling species distributions? J Appl Ecol 41: 193–200. Google Scholar

82.

Russell IA, Randall RM & Hanekom N (2014) Spatial and Temporal Patterns of Waterbird Assemblages in the Wilderness Lakes Complex, South Africa. Waterbirds 37: 1–18. Google Scholar

83.

Santillan V, Quitian M, Tinoco BA, Zarate E, Schleuning M, Bohning-Gaese K et al. (2018) Spatio-temporal variation in bird assemblages is associated with fluctuations in temperature and precipitation along a tropical elevation gradient. PLoS ONE 13: e0196179  https://doi.org/10.1371/journal.pone.0196179  Google Scholar

84.

Sathe SS, Khabade S & Hujare M (2001) Hydrobio-logical studies on two manmade reservoirs from Tasgaontahashil (Maharashtra). Ecol Environ and Conserv7: 211 – 217. Google Scholar

85.

Singh RP, Mathur P (2005) Investigation of physicochemical characteristics of freshwater reservoir of Ajmer city Rajasthan. Indian J Environ Sci 19: 179–186. Google Scholar

86.

Stockwell DRB & Peters D (1999) The GARP modelling system: Problems and solutions to automated spatial prediction. Intl J Geogr Inf Sci13: 143–158. Google Scholar

87.

Sulai P, Nurhidayu S, Aziz N, Zakaria M, Barclay H, & Azhar B (2015) Effects of water quality in oil palm production landscapes on tropical waterbirds in Peninsular Malaysia. Ecol Res 30: 941–949. Google Scholar

88.

Tang ZY, Fang JY, Sun JY & Gaston KJ (2011) Effectiveness of protected areas in maintaining plant production. PLoS ONE 6: e19116.  https://doi.org/10.1371/journal.pone.0019116  Google Scholar

89.

Tank SK, Chippa RC (2013) Analysis of water quality of Halena block in Bharatpur area. Intl J Sci Res 3: 1–6. Google Scholar

90.

Thapa JB & Saund TB (2012) Water Quality Parameters and Bird Diversity in Jagdishpur Reservoir, Nepal. Nepal J Sci Technol 13: 143–155. Google Scholar

91.

Thuiller W (2003) BIOMOD: optimising predictions of species distributions and projecting potential future shifts under global change. Global Change Biol 9: 1353–1362. Google Scholar

92.

Thuiller W, Lafourcade B, Engler R & Araujo MB (2009) BIOMOD: a platform for ensemble forecasting of species distributions. Ecography 32: 369–373. Google Scholar

93.

Tian B, Zhou YX, Zhang LQ & Yuan L (2008) Analysing the habitat suitability for migratory birds at the Chongming Dongtan Nature Reserve in Shanghai, China. Estuar Coast Shelf S 80: 296–302. Google Scholar

94.

Tucker K, Rushton SR, Sanderson RA, Martin EB & Blaiklock, J (1997) Modelling bird distributions – a combined GIS and Bayesian rule-based approach. J Landsc Ecol 12: 77–93. Google Scholar

95.

Vallecillo S, Maes J, Polce C & Lavalle C (2016) A habitat quality indicator for common birds in Europe based on species distribution models. Ecol Indic 69: 488–499. Google Scholar

96.

Van Niekerk JH (2010) Assemblages and movements of waterfowl at cattle feedlots across Gauteng, South Africa. Ostrich 81: 31–37. Google Scholar

97.

Vanagas G (2004) Receiver operating characteristic curves and comparison of cardiac surgery risk stratification systems. Interact. Cardiovasc Thorac Surg 3: 319–322. Google Scholar

98.

Vapnik V & Izmailov R (2018) Rethinking statistical learning theory: learning using statistical invariants. Mach Learn 108: 381–423. Google Scholar

99.

Wainger L & Mazzotta M (2011) Realizing the potential of ecosystem services: a framework for relating ecological changes to economic benefits. Environ Manage 48: 710–733. Google Scholar

100.

Weller MW (1999) Wetland Birds: Habitat Resources and Conservation Implications. p 316. Cambridge University Press, New York. Google Scholar

101.

Wetlands International (2018) Available atIWC Online database. URL: http://iwc.wetlands.org(accessed on 29 October 2018). Google Scholar

102.

World Bank Report (2011) Bird species; threatened in Malaysia. Available at URL: http://www.tradingeconomics.com/Malaysia/birdspecies-threatened-wbdata.html(accessed on 2 January 2013). Google Scholar

103.

Wormworth J & Mallon K (2014) Bird Species and Climate Change: The Global Status Report version 1.0. A synthesis of current scientific understanding of anthropogenic climate change impacts on global bird species now, and projected future effects. p 75. World Wide Fund for Nature, Canberra. Google Scholar

104.

Yang H (2013) The case for being automatic: Introducing the automatic linear modelling (LINEAR) procedure in SPSS statistics. Mult Linear Regr View39: 1–37. Google Scholar

105.

Zainul-Abidin M K, Mohd-Taib F R & Md-Nor S (2017) Distribution and habitat selection of the Asian Openbill (Anastomusoscitans) in Peninsular Malaysia. Mal Nat J 69: 169–181. Google Scholar

106.

Zakaria M & Rajpar MN (2010) Density and diversity of waterbirds and terrestrial birds at Paya Indah Wetland Reserve, Selangor Peninsular Malaysia. J Biol Sci 10: 658–666. Google Scholar

107.

Zakaria M & Rajpar MN (2013) Density and Diversity of Water Birds and Terrestrial Birds in Man-made Marsh, Malaysia. Sains Malays 42: 1483–1492. Google Scholar

Appendices

Appendix

Image preprocessing and data conversion

The Sentinel 2A MSI bands contained top of Atmosphere reflectance integer units. Therefore, each band was converted to a floating point and then divided by 10000 prior to analysis using the raster calculation tool of ArcGIS 10.4. Then, they were subjected togeometric and radiometric corrections with the aid of the annotated ancillary information embedded in the Sentinel 2A dataset's support files. The spatial reference systems of the wetlands' satellite datasets and vector data were changed (coordinate transformation and datum projection) from World Geodetic System 1984 (WGS84) to the Malaysian local projected coordinate system (Selangor GDM 2000). Radiometric correction was performed on the transformed raster and vector datasets using histogram equalization, haze and noise reduction functions in ERDAS Imaging 2014 software. The image enhancement is to improve its features' visual display without causing any spectral distortion. Furthermore, the presence/background, climatic and hydrology data were converted to delimited text format in Microsoft Office 2007, imported into the ArcGIS 10.4 environment and converted into respective shapefiles.

© The Ornithological Society of Japan 2021
Chukwuemeka Onwuka Martins, Oluwatobi Emmanuel Olaniyi, and Mohamed Zakaria "Environmental Factors and Spatial Heterogeneity Affect Occupancy Estimates of Waterbirds in Peninsular Malaysia," Ornithological Science 20(1), 39-55, (7 April 2021). https://doi.org/10.2326/osj.20.39
Received: 19 June 2019; Accepted: 20 May 2020; Published: 7 April 2021
JOURNAL ARTICLE
17 PAGES


SHARE
ARTICLE IMPACT
Back to Top