Heronry distribution and site preference dynamics of tree-nesting colonial waterbirds in Tamil Nadu

Study area

The study was conducted in Tamil Nadu, which is broadly divided into three physiographic divisions: the north-western high-elevation regions, the western hilly area and the plains along the coast (Subramanya, 2005). The climate is semi-arid in the plains and humid to sub-humid in the hills. For administrative purposes, the state is divided into 32 districts and seven agro-climatic zones (Fig. 1). Whereas the rest of the country receives most of its rainfall during the south-west monsoon (SWM), i.e., from June to September, Tamil Nadu gets its rainfall during the north-east monsoon (NEM), i.e., from October to December. The state of Tamil Nadu lies in the rain shadow region of the Western Ghats due to which it receives limited rainfall during the SWM (June–September). The average annual rainfall of the state is about 921 mm and is largely dependent on the NEM to replenish its bodies of water (Subramanya, 2005; Singh et al., 2019). Both the livelihoods of the local people and the breeding cycles of the colonial nesting waterbird species are affected, if there is a monsoon failure (Jabaraj & Gopi, 2020).

The state has 17 river basins (Palanisami et al., 2011) and about 7% of the geographic area of Tamil Nadu is comprised of wetlands (rivers and reservoirs) (SAC, 2011). A total of 24,684 inland wetlands have been mapped at a scale of 1:50,000 in Tamil Nadu (SAC, 2011). The major wetland types in the state are rivers, lakes, reservoirs, ponds and tanks (SAC, 2011). Tanks are small and have mainly been constructed for storing water for household and irrigation purposes (Subramanya, 2005), whereas reservoirs include dams and lakes and are larger in size. The tanks in India are considered as the largest source of irrigation and they are highly concentrated in the southern states.

The tanks across the state, especially those near dams, store water during the lean season and supply water for agricultural purposes, mainly rice cultivation (Von Oppen & Rao, 1980). These tanks, along with the rivers, reservoirs and coastal wetlands, serve as excellent feeding, roosting and nesting grounds for colonial nesting waterbirds and as stopovers for migratory birds. This study was carried out during the period from 2017–2019. The two-study period was defined as the first year (October-2017 to September-2018) and the second year (October-2018 to September-2019). The survey was conducted predominantly at each district with further focus at the taluk (subdivision of a district) level. Initially, literature was studied to identify already existing heronries. A Survey of India map (1:50,000) was additionally used. Waterbodies were located using Google Earth and were surveyed for the presence of breeding colonies of colonial nesting waterbirds. A total of 123,486 km was travelled across the state during the study period. The information about the congregation of waterbirds was obtained from eBird (2017). The XXVII meeting of the Training, Research and Academic Council (TRAC) of the Wildlife Institute of India approved this work (21-02-2012). The Principal Chief Conservator of Forests and Chief Wildlife Warden approved a permit to enter protected areas for this research (Ref No. WL5(A)/19603, Permission No. 32/2017, Dated 07-07-2017).

Waterbird survey

Waterbirds in the breeding colonies were identified and counted, and information pertaining to the GPS location of the heronries, number of species nesting, details of the location and nesting trees was documented in the prescribed format. When counting colonial nesting waterbirds, it is most important to consider the fluctuations in the number of adult birds during the day. Colonial nesting waterbirds spend a significant part of the time outside the nesting area, either for foraging or collecting nesting material. Hence, the count must be carried out at the same time of each day at a particular site. The standard counting protocol used for colonial nesting waterbirds, i.e., total counts, was followed at each nesting site (Bibby et al., 2000). A vantage point was selected at each site, and the count was carried out in the morning before the birds had left the heronry (between 0500 and 0700 h) and repeated in the evening (between 1600 and 1800 h).

One of the major concerns was the detectability of the waterbird colonies. The survey involved a large geographical area, hence there was a high probability of missing breeding locations. To address this issue, local birdwatchers and non-governmental organizations (NGOs) assisted with locating known waterbird nesting colonies. One of the main objectives of the study was to locate previously unknown colonies and maximum effort and time was invested in reporting potential waterbird nesting colonies across the landscape. Two major challenges were the probability of missing a nesting colony if the area was inaccessible (e.g., aquatic habitats) or that the timing of the survey was incorrect for the target species. This was overcome by using country boats to survey waterways and aquatic habitats that were not accessible by roads and conducting the survey during both the monsoons (SWM and NEM). Moreover, standard protocols pertaining to colonial waterbirds survey were followed (Steinkamp et al., 2003; Cavitt et al., 2014).

Species distribution modelling

The suitability of sites for heronries was modelled using MaxEnt (Phillips, Anderson & Schapire, 2006). The global climate database WorldClim was used to acquire climate-related data (www.worldclim.org). Cumulative temporal weather data (1950–2000) was used to calculate the mean annual temperature, and a raster of resolution one km was extracted. The average precipitation (in millimetres) during the peak breeding season of the colonial nesting waterbirds, elevation (in metres), forest cover, land use/land cover and night-time light data (one-km resolution) were obtained (Table S1). The precipitation, temperature, forest cover and elevation were used as covariates as there was no pattern in the land use/land cover and night-time light data. To ensure better performance with the smaller data set, fivefold cross-validation was used, with the area under the curve (AUC), to assess the model. The regularization was kept at the default value of 1.0 to avoid over-fitting. The best-fitting model was chosen using the jack-knife variable selection procedure.

Data analyses

The relative abundances of the colonial nesting waterbirds were calculated for the two survey periods. The relative abundance (R) of waterbirds was calculated using the formula

R = (Ni/Tn) ∗ 100

where,

Ni = total number of individuals of the ith species

Tn = total number of individuals of all species

The species accumulation curve was used to evaluate the relationship between the sampling effort and species detection over the survey period. Moran’s I index was used with a Euclidian distance estimator to check whether the sites were clustered, random or dispersed (Getis, 2008). All the statistical analyses were performed using the statistical program R (R Core Development Team, 2019).

Distribution of heronries

The highest number of heronries were recorded in Tirunelveli, Ramanathapuram and Kancheepuram districts, with 12 heronries in each as per the previous study (Subramanya, 2005). As per Ganesh, Abhisheka & Prashanth (2014), Tirunelveli and Thoothukudi accounted for ten heronries. In the present study, Ramanathapuram district accounted for only six active heronries. Eight districts had only one heronry each and heronries were not recorded in ten districts of the state during the study period. Active nesting sites were recorded in Salem and Karur districts, which were devoid of heronries in earlier studies. More than 70% of heronries, that were active for the past 3–5 years, were reported for the first time in the present study. Three sites, Madras Crocodile Bank Trust, Dhamal and Samantham tank, that were not active during the earlier study were found to host colonial nesting waterbirds during the current study.

An earlier study, which also included erstwhile heronries, reported 97 heronries, but only 58 active nesting sites were recorded during that study (Subramanya, 2005). A total of 15 sites were common to the present study which included 101 sites and the earlier study which included 58 active sites. Continuous monitoring is required to understand the relationship between colonial nesting waterbirds and the environmental factors and habitat quality, particularly at the nesting areas (Ismail & Rahman, 2012). The biggest challenge in addressing the occupancy of waterbird species and their population dynamics is the lack of studies on landscape effects, especially in Tamil Nadu. For species that often change breeding sites, conservation and management interventions should be taken at a landscape level rather than at specific sites (Wyman, Wires & Cuthbert, 2014). As these birds are subjected to variety of threats, a landscape level approach in terms of monitoring and surveys should be regularly conducted during the peak breeding season across the region.

Nesting pattern of colonial nesting waterbirds

The Little Cormorant was recorded in about 80 heronries, both in protected and non-protected areas. It was also found outside wetlands, especially along roadsides, as a single-nesting species or with two to four species. The Little Egret was found at 67 nesting sites, the Black-crowned Night Heron at 52 and the Indian Pond Heron at 44. The Oriental Darter nested at 29 sites, which is significantly higher than previously recorded. The Spot-billed Pelican nested at 16 sites and the Black-headed Ibis at 19 sites. The Indian Black Ibis was observed as the least common nesting species in Tamil Nadu (Table S5). The Great Cormorant was next, nesting in five heronries. Our low detection of species such as the Cattle Egret and Indian Black Ibis could be due to the fact that they normally prefer to nest away from mixed-species heronries in this landscape, though, the Cattle Egret is known to breed in monospecific and mixed colonies (Dwevedi, Kumar & Mylswamy, 2015). Species such as the Painted Stork, Spot-billed Pelican, Black-headed Ibis, Asian Openbill, Eurasian Spoonbill and Great Cormorant nested at well protected sites. Except for two sites, the Painted Stork nested predominantly in bird sanctuaries. Colonies of this species as well as the Spot-billed Pelican were found only at sites that enjoyed the utmost protection. In most of the heronries, the nesting of the Spot-billed Pelican began before the Painted Stork contrary to the previous study (Subramanya, 2005). Koonthankulam is known to host the Spot-billed Pelican; the oldest record goes back many centuries (Webb-Peploe, 1945). These reports indicate that the Spot-billed Pelican has nested in the southern states for more than a century and that these states have become a traditional home for this species (Subramanya & Manu, 1996; Subramanya, 1996, 2005). A total of 11 of the 17 nesting sites of the Asian Openbill were in the districts on the east coast, and six nesting sites were in the inland districts of Tamil Nadu.

The Great Cormorant has been reported to nest around reservoirs in the Nilgiris District, Koonthankulam Bird Sanctuary and Vedanthangal Bird Sanctuary (Subramanya, 2005), but nesting was not observed at either of these two bird sanctuaries in the present study. Active nesting sites were observed at three bird sanctuaries, Vaduvoor, Karaivetti, Vellode, and at one site each in the Nilgiris and Krishnagiri Districts. All the three ibis’ species (Black-headed Ibis, Indian Black Ibis and Glossy Ibis) breed in Tamil Nadu (Subramanya, 2005); the Black-headed Ibis is the most common of the three ibis’ species, followed by the Glossy Ibis. Interestingly, the Indian Black Ibis was never found nesting along with other species during the present study and was only observed breeding in Tirunelveli District. This species was first recorded nesting in Koonthankulam Bird Sanctuary (Wilkinson, 1961). From a conservation perspective, for species that nest away from wetlands, it is important to protect their traditional sites. For example, in the south-eastern coastal districts of Tamil Nadu, the Indian Black Ibis prefers the palmyra palm for building its nests. Both single-species and mixed-species heronries occur in Tamil Nadu. Two heronries had 16 species nesting—these were the oldest bird sanctuaries in Tamil Nadu (Vedanthangal and Koonthankulam). Ten heronries had 11 to 15 nesting species, and 18 sites had six to ten species. The majority (71 heronries) had one to five species, with around 16 of them being single species nesting colonies (Table S6). Sites that have large numbers of nesting species tend to be in bird sanctuaries or in areas where people actively protect the nests during the breeding season (e.g., Vagaikulam heronry) (Abhisheka et al., 2012).

Nesting tree species

The habitat selection and population dynamics of colonial nesting waterbirds are influenced by various factors, among which nesting tree quality is one of the predominant criteria (Baxter & Fairweather, 1998; Roshnath & Sinu, 2017). Vachellia nilotica was found in most of the wetlands planted by the forest department in the early 1960s to meet the demand for fuelwood (Wilson, 1979). Most of the tanks in Tamil Nadu were planted with this species, which resulted in thick vegetation. Vachellia nilotica attracted waterbirds in large numbers and supported nesting activities. Before this species was planted, colony-nesting waterbirds built their nests amidst human habitations. They gradually shifted to the Vachellia nilotica plantations in the tanks (Subramanya, 2005). One example is Koonthankulam, where waterbirds initially preferred trees amidst human habitations and later shifted to the trees in the tanks.

Trees located along roadsides (national highways/state highways), near residential areas and in non-residential areas (inside campuses) have been recorded as excellent nesting habitats for colonial nesting waterbirds (Subramanya, 1996; Sashikumar & Jayarajan, 2007; Roshnath & Sinu, 2017). The present study corroborates this. Tamarindus indica and Ficus religiosa trees growing along roadsides, near reservoirs, along the banks of large rivers and in temple premises, provide excellent nesting habitats for the Little Cormorant, Little Egret, Indian Pond Heron and Black-crowned Night Heron. Tree species such as Peltophorum pterocarpum and Delonix regia, especially if located in campuses, provide conditions conducive to nesting. In general, many institutions and factories provide shelter to colonial nesting waterbirds. Avicennia marina was a mangrove species that was used by waterbirds extensively. Artificial nesting structures (Jabaraj & Gopi, 2021) have been constructed in recent years and are extensively utilized by colonial nesting waterbirds.

Protected vs. non-protected heronries

One of the most effective measures employed in conservation is the establishment of legally protected areas (Dudley, 2008). The protected area network of Tamil Nadu includes bird sanctuaries, conservation reserves (areas owned by state governments adjacent to protected areas (PAs), for protecting the landscape, seascape and habitat of fauna and flora), reserve forests and islands. The species richness and diversity of waterbirds in the bird sanctuaries were higher in comparison with the rest of the protected area network (conservation reserves, reserve forests). Tamil Nadu has 15 bird sanctuaries and two conservation reserves. Colonial nesting waterbirds nest in nine of the bird sanctuaries and the two conservation reserves. During the study there was no nesting at four bird sanctuaries: Chitrangudi, Kanjirakulam, Sakkarakotai (Ramanathapuram District) and Karikili (Kancheepuram). A total of 82 heronries are situated outside protected areas and this is the biggest challenge for conservation of these heronries. Most areas with rich avifaunal biodiversity lie outside the legally protected areas. Urbanization is one factor that has changed the availability of nesting habitats (Lim & Sodhi, 2004; Urfi, 2010). Most of the nesting sites are dispersed within wetlands, along roadsides or in institutions and factories with no protection. Some of the sites located outside protected area networks enjoy protection through the efforts of local people (e.g., Vagaikulam heronry), the authorities of the factories (e.g., TVS motors), educational institutions and religious groups (Jabaraj & Gopi, 2021). Around 21 heronries fall in the non-legally protected area category (protected by people). There is minimal disturbance to the nesting colonial waterbirds, especially during the breeding season, and they are protected from hunting. In both years of the survey, the species abundance and richness were similar in the protected areas and the non-protected areas unless there was a failure of the monsoon and a dearth of water in a wetland (Jabaraj & Gopi, 2020). Waterbirds nesting inside campuses and religious premises are safeguarded from external disturbance; one good example is the Melmaruvathur heronry, where the colonial nesting waterbirds are given the utmost protection by the temple administration (Jabaraj & Gopi, 2021). Partnership between the local people and concerned authorities is critical in protecting the existing heronries located outside protected areas.

MaxEnt: species distribution modelling

Breeding behaviour shaped by the climatic and anthropogenic changes in the landscape (Huntley et al., 2006) can be predicted using SDMs (Peterson, 2001). The MaxEnt species distribution model developed, predicted breeding habitats for colonial nesting waterbirds in Tamil Nadu. Presence-only data was used for the model; the predicted sites mostly clustered near waterbodies, along rivers and reservoirs and in areas where dense vegetation is close to waterbodies. The site-level variables required to understand the nesting ecology of waterbirds have still not been studied in depth (Kreakie, Fan & Keitt, 2012); parameters that could influence nesting include water depth, disturbance level, vegetation, distance from roads and settlements and water quality etc. (Maleki et al., 2016; Sheehan et al., 2017). Other landscape complexities that may also influence the nesting sites of these birds need to be identified and considered in future models.

The predicted hotspots were strongly corroborated (AUC: 0.75) by the field survey data. The wetlands in the high-probability areas were not strictly nesting sites but were used either as feeding or roosting sites, which can later be converted to nesting sites. Gurusami (1994) also observed this type of site behaviour. Environmental characteristics can be used to model the distribution of wildlife, and such modelling can help develop effective management plans to prioritize areas that may suffer major changes in a changing landscape (May, Page & Fleming, 2016). Due to the dynamic nature of the wetlands and our poor understanding of the impacts of climate change, it is hard to foretell the response of wetland species, especially waterbirds, to the ongoing landscape modifications and climate change. Moreover, climate change, which directly or indirectly influences the monsoon pattern, has a direct impact on the breeding colonies in a particular year (season), as is very evident from the present study. The breeding and the population are affected if the monsoon fails (Jabaraj & Gopi, 2020). Some of the sites located in human-dominated landscapes are completely at the mercy of the local administration or the people; in some cases, the wetlands are used for developmental activities (Panigrahy et al., 2012). To create better predictive models for larger landscapes further site-level studies must be carried out to understand the environmental variables that might influence the nesting behaviour of these birds.

Threats

Habitat destruction (Fig. S1) and hunting (Fig. S2) remain the biggest threats to colonial nesting waterbirds. The novel techniques used by hunters targeting agricultural fields have become a major concern in recent years. Disturbance in the form of fishing (Fig. S3), collection of firewood and livestock grazing within the wetland affects the waterbirds nesting during breeding season. About 50% of the heronries recorded in the state occur outside wetlands among densely populated human habitation and along roadsides. Major threats to waterbird breeding include the felling of nesting trees during road widening, local residents reacting to the nuisance created by waterbirds, the collection of eggs from the nests and hunting. Species like Egrets, Herons and Cormorants are hunted across the state. Further encroachments within the wetlands and extensive guano deposition causing nesting tree death (Fig. S4) are other threats that must also be evaluated. The attitude of the people towards heronries also remains an important determining factor for the persistence of these species in this landscape. People occasionally drive away the birds before they start nesting by bursting crackers and other methods as they find guano excretion, noise, falling of the eggs and dead chicks to be a nuisance. Along with the above threats, climate change remains an important factor influencing these colonial nesting waterbirds (Cavitt et al., 2014).