Diversity and abundance of aquatic birds in Koonthankulam village pond, Tamil Nadu, India

: The diversity of birds in Koonthankulam pond, located in Koonthankulam village (8.495N, 77.755 E), Tirunelveli district, southern Tamil Nadu, was studied. A total of 90 species belonging to 21 orders, 42 families, and 73 genera were recorded. The study recorded seasonal migrants such as Black Ibis, Oriental White Ibis, Bar-headed Goose & Spoonbill and indigenous species including the Pond Heron, Cattle Egret, White-breasted Kingfisher, Red-wattled Lapwing, Rose-ringed Parakeet, Purple-rumped Sunbird, Hoopoe, and Indian Robin. The primary data were analyzed by principal component analysis, cluster, and analysis of variance. Analysis of variance showed that the Menhinick index is statistically significant P <0.05. A structural equation model was applied to analyze the physico-chemical parameters of water samples collected from the sampling site. Analysis of experimental data through the structural equation model indicates temperature and dissolved oxygen may indirectly affect bird diversity.


INTRODUCTION
The process of urbanization has fragmented and degraded different types of habitats. One such habitat is ponds of varied sizes, especially in urban and semiurban areas. Under such conditions, the existing ponds provide little hope for life and support for the survival of organisms. Wetlands are among the most productive ecosystems in the world and play vital roles in flood control, aquifer recharge, nutrient absorption, and erosion control. In addition, wetlands provide home for a huge diversity of wildlife such as birds, mammals, fish, frogs, insects, and plants (Buckton 2007). Among several organisms surviving in and around water bodies, birds occupy a significant position, as they are one of the critical ecosystem functionaries.
Birds play prominent roles in ecosystems, serving as pollinators (Stiles 1978), predators (Rudebeck 1950), scavengers (Roen 2005), prey (Rudebeck 1950), and regulators of pest populations (Peterson 1980). Their interactions are wide and varied with abiotic and biotic components of different ecosystems, i.e., they are not restricted to one particular system but also to adjacent systems as they enjoy the power of flight. India hosts around 1,353 species of birds (Praveen & Jaypal 2023). Analysis of avian diversity portrays the status of their aquatic habitats and neighboring ecosystems. As there is no detailed report on the diversity of Koonthankulam village pond, the present study was carried out to analyze seasonal variation in bird diversity and their relationship with water quality parameters.

MATERIALS AND METHODS
The study was carried out in Koonthankulam village pond (8.495N, 77.755E), Tirunelveli, southern Tamil Nadu, from January 2017 to November 2018. This pond is surrounded by agricultural fields, where different crops are grown throughout the year. Macro-invertebrates of the agricultural fields and grains scattered around after harvesting along with the pond allure avifauna to this region. Bird watching and recording have been carried out for six seasons (namely, spring, summer, early monsoon, late monsoon, early winter, and late winter) by point count protocol as per Newson et al. (2009). Observations were made using a binocular (Nikon 16x50 AculonA211), and photography was done with Canon 6D Mark II with zoom lenses. The birds recorded were identified by referring to Ali & Ripley (1981).
Physical and chemical parameters such as temperature, pH, total dissolved solids (TDS), conductivity, salinity, and dissolved oxygen (DO), were measured on the spot using a water analyzer (Systronic make 371). Other parameters (Hardness, magnesium, calcium, chloride, alkalinity, and acidity) were determined following the standard procedure from American Public Health Association (APHA) and Trivedy & Goel (1984). The map has been generated using the software QGIS 3.6. Structural equation modeling (SEM) is a multivariate statistical tool that can be used to describe linear relationships among variables (McCune & Grace 2002;Grace 2006). SEM provides explicit regression estimations for all parameters (Byrne 2001). Structural equation modeling of groundwater physicochemical parameters data was used to characterize the groundwater quality and to identify the controlling factors on bird diversity. IBM SPSS AMOS 22.0 was used to analyze the structured model's fit and estimate the parameters of both observed and latent variables. Chi-square test, the root mean square error of approximation (RMSEA), and the goodness fit index are used as measures of model fit. A measure of minimum sample discrepancy is indicated by the value chi-square divided by the degrees of freedom (CMIN/df) (Belkhiri & Narany 2015). This measure was used to analyze the fit of the model. A value of less than 5 indicates the model's fit is adequate (Arbuckle 2012), less than 3 reflects that the model is acceptable (Kline 1998), whereas a value of 2 or less represents the model was fit as a good model. Goodness-of-Fit statistics (GFI) was calculated as the variance proportion accounted for by the estimated covariance (Tabachnick & Fidell 2007). The RMSEA provides a way to understand optimally chosen parameter estimates that would fit the covariance matrix (Byrne 1998). When the proposed structural model has a (comparative fit index) CFI>0.95 and an RMSEA <0.05, then the structural model is to be considered a good model (Byrne 2010). The diversity indices were calculated using PAST 3.14 software.

RESULT AND DISCUSSION
A total of 90 species of birds belonging to 21 orders and 42 families were recorded. Of this, 87 species of birds are of 'Least Concern', and three species are 'Near Threatened' (Table 1). Of these, 41 were waterbirds, and 49 were terrestrial birds (Table 1). Waterbirds constitute 12 species of waterfowl (swimmers), 25 species of waders, and three species were divers. Out of 49 terrestrial birds, 35 species were passerine birds, five were birds of prey, J TT six were upland ground birds, two were night birds, and two were tree-clinging birds. It is evident from the data that the order Passeriformes is represented by the most families (Sturnidae, Motacillidae, Corvidae, Dicruridae, Estrildidae, Muscicapidae, Alaudidae, Motacillidae, Nectariniidae, Passeridae, Pycnonotidae, Muscicapidae, Monarchidae, and Leiothrichidae). In contrast, the highest numbers of species recorded were from Ardeidae, Anatidae, Cuculidae, Columbidae, Rallidae, and Threskiornithidae. The highest abundance index (9.2) was seen for Anastomus oscitans. Bubo bubo and Pandion haliaetus had a low abundance index (0.02). Of the 90 species, 26 species have more than 1 abundance index.
The similarity in the species composition and abundance among the six seasons analyzed by Bray-Curtis coefficient (Cluster analysis) clustered the

J TT
seasons into five, in the range of 55.25-100.00 ( Figure  3). The five clusters show that each season has a different composition of the bird populations. The dendrogram showed that summer and early monsoon have a maximum similarity of 76.30. Two groups were identified among the six seasons. Early winter, late winter, and spring formed a group and early monsoon, late monsoon, and summer formed another group. The number of individuals across seasons differed significantly (ANOVA, F 5,12 = 6.673, P <0.05; Table 2). A higher number of individuals was present in late winter (3027.33 ± 498.27), whereas the lowest number was recorded in late monsoon (1315.00 ± 158.24). The second-highest population of birds appeared in spring (2980.33 ± 316.26). This implies that the number of birds from December─March was high. The maximum Taxa_S was found in late winter (85.00 ± 1.73), whereas the minimum was in late monsoon (80.66 ± 0.33), with the range of Taxa_S over all the seasons being 5. The results reveal minimum deviation in species composition with high variation in population. However, the highest Shannon_H diversity (3.70 ± 0.06) was in late monsoon and the lowest (3.36 ± 0.19) in spring, indicating a more diverse and even species distribution in late monsoon. The Shannon_H diversity of birds among various seasons was not significantly different (ANOVA, F 5, 12 = 0.898, P >0.05).
The species evenness among the various seasons was measured by Buzas and Gibson's index. Evenness was maximum in late monsoon (0.50 ± 0.03) and minimum in spring (0.35 ± 0.06). However, the Buzas and Gibson's evenness indices across various seasons were not significantly different (ANOVA, F 5, 12 = 1.24, P >0.05). The richness was measured by the Menhinick species richness index. The Menhinick richness index differed significantly among the seasons (ANOVA, F 5, 12 = 7.811, P <0.05). Early (2.25 ± 0.12) and late monsoon (2.25 ± 0.13) have a high value of Menhinick richness index, and spring (1.56 ± 0.12) and late winter (1.57 ± 0.09) have a low value. The Chao-1 estimator was used to analyze singleton and doubleton species in the bird community. The maximum singleton and doubleton species occurred   in summer (101.27 ± 9.26) and the minimum in late monsoon (90.89 ± 0.37). Among all the seasons, the Chao-1 estimator was not significantly different (ANOVA, F 5,12 = 0.518, P >0.05).
A correlation analysis has been carried out for all physico-chemical parameters with swimmer, diver and wader species to understand the influence of the waterquality parameters on the bird population (Table 3). The following results were noticed in that analysis. The temperature was negatively correlated with swimmer and wader populations at p = 0.01, and DO was positively correlated with all the three water bird communities encountered in this study at p = 0.05. In addition to this, a SEM model was designed and analyzed to confirm that temperature and DO were the most important parameters that affect bird counts. Thus, a structural equation model has been framed with the following six parameters, temperature, DO, swimmer, diver, and wader abundances. Bird's abundance has been studied with reference to the effect of temperature using structural equation modeling, as in Duclos et al. (2017). Patra et al. (2010) used stepwise multiple regression analysis to study the physico-chemical parameters affecting the avifaunal abundance; e1, e2, and e3 are added to the SEM to reduce the error value between the variables. Temperature, DO, and errors are exogenous variables, and swimmer, diver, and wader birds are endogenous variables.