Effect of ecological factors on the grass dynamics at Point Calimere Wildlife Sanctuary, India

: Grass dynamics play a major role in the density and diversity of grazing mammals. To understand the drivers of grass quality and quantity, we assessed the height, cover, soft-texture, green leaves, and reproductive phase of grass species in relation to 13 ecological covariates belonging to climate, vegetation, human disturbance


INTRODUCTION
In grassland environment grass phenology regulates the life cycles of grasses, which has a direct impact on biodiversity and trophic levels associated to herbivory (Auken 2009;Fischer et al. 2013).Phenology of grass species such as fledgling grass, soft shoots, soft, and green leaves determine herbivore nutrients (Hughes et al. 1993).Leaf length, leaf size, and spruce growth are unique features of each species (Ahsan et al. 2010;Huijser & Schmid 2011;Wang et al. 2011).The growth rate depends on the life cycle of the grass species.At the vegetative phase, cellulose and hemicellulose are present in large quantities for energy supply (Islam et al. 2003;Hussain & Durrani 2009).The budding, flowering, fruiting to seed dispersal stages are the reproductive phases (Sherry et al. 2011).Many factors, including rainfall, soil type, season, access to water, and the availability of open habitat play a role in determining the quantity and quality of grass that can grow (Ganskopp & Bohnert 2001;Sawyer et al. 2005;Hagenah et al. 2008;Hussain & Durrani 2009;Zeng et al. 2010).
Grasslands are significant global reservoirs of biodiversity and important food sources for herbivores (Jing et al. 2014;White et al. 2000;Bardgett et al. 2021).Natural grasslands are also vital to climate and water regulation, and to biogeochemical cycles like carbon balance, hence their degradation has serious consequences (White et al. 2000;O'Mara 2012;Cai et al. 2015).Natural grasslands cover regions with sufficient precipitation for grass to grow.Climatic, human-caused, and other environmental factors influence grasslands and alter grass phenology (Boval & Dixon 2012).Worldwide, grasslands have been disappearing for the greater part of a century (Egoh et al. 2016).In a short amount of time, grassland can be negatively affected by a shift in land use.Specifically, a major issue with grasslands is the growth and succession of forests (Liu et al. 2013).
This study focuses on the grass species of Point Calimere Wildlife Sanctuary, which acts as the major food source to the Blackbuck, an iconic species of the reserve, and other ungulates.Multiple pressures, from the invasion of Prosopis to over-grazing by cattle and feral horse on the grasslands habitat, would result in decline in grass biomass needed to support the Blackbuck population in the study area (Baskaran et al. 2016;Arandhara et al. 2020Arandhara et al. , 2021)).Although grasses have wide ecological amplitude and several adaptations to withstand trampling, grazing, fire, food, and drought, they face severe competition for light and nutrients from aggressive wood species and invasive plants in tropical forests (Ashokkumar et al. 2021).This study assessed grass species parameters representing quantity: height & cover, and quality: grass soft-texture, green grass, & reproductive phase, across three habitats and two seasons between 2018─2020 to identify the ecological drivers of grass quantity and quality.

Study area
The research was carried out at the Point Calimere Wildlife Sanctuary in Tamil Nadu, which is located between 10.30-79.85 0N and 10.35-79.42 0E at the confluence of the Bay of Bengal and the Palk Strait, near Nagapattinam (Figure 1).The reserve encompasses 30 km 2 of dry evergreen forest, grassland, open scrub, sandy coastline, salt marshes, and backwaters (Ali 2005).The grasses in the sanctuary's southernmost region cover 17% (4.49km 2 ) of the total sanctuary area.The sanctuary's native and flagship species of Blackbuck live in grasslands that also serve as a foraging ground for other herbivores like feral horses, Chital, and domestic cattle.The average annual rainfall in Point Calimere is 1,366 mm, with temperatures ranging 23─37 ºC.The grasslands are especially vulnerable to invasion by Prosopis juliflora.Anthropogenic pressures on the sanctuary include firewood collection, fishing, and cattle grazing.

Data collection
Assessment of Grass dynamics and other variables: Data were collected between November 2018 and October 2020 covering two seasons (dry season: February─August and wet season: September─January) and three habitats (Dry evergreen, open scrub, and grasslands).The study area was overlaid with 1 km 2 grids and placed with a 1-km line transect at each grid.The grass availability and parameters were evaluated on a monthly interval.Four 1 m 2 quadrates were placed at 5 m intervals on the north, south, east, and west directions at every 250 m interval along these transects.Quadrates of this size have previously been widely used in studies of grass abundance (Menut & Ceaser 1979;Hacker 1984;Sivaganesan 1991).In total, 1,024 quadrates were laid (dry evergreen-453, open scrub-272, and grasslands-299) (Image 1).All grass dynamic parameters were recorded in each quadrat following methods described in Table 1 Grass specimens were collected and preserved in order to create herbariums for each grass species for species identification and confirmation (Rangel et al. 1999;Shaw 2008;Shankar & Shashikala 2010).The vouchered

Data analysis
Statistical analysis: Prior to performing a detailed analysis, the compiled data were examined for normality and variance homogeneity.The Kolmogorov-Smirnov (KS) test used on grass parameter such as, height (KS: 0.35; p = <0.05),cover (KS: 0.19; p = <0.05),soft texture (KS: 0.23; p = <0.05),green leaves (KS: 0.16; p = <0.05),and reproductive phase (KS: 0.37; p = <0.05) of grass species was neither normal, nor could be transformed to normal with four different transformations (Arsin, Log10 (LG10), Inverse log, and Exponential).Therefore, the difference in the selection of this species between seasons were tested using non-parametric Mann-Whitney U-test and Kruskal-Wallis H test. SPSS V.23 software package was used for the statistical test and the general linear model (GLM) was used to identify the covariates influencing on grass dynamics characteristics and all parameter models were ranked by their small-sample Akaike information criterion (AICc) and inferences were taken from models with ΔAICc ≤ 2. However, in results are comprehensively shows that top two models of lowest ΔAICc.Model comparison were calculated by the R package 'MuMIn' (Barton et al. 2018) by in R Library, in R Software Version 3.3.3(R Core Team 2019).This model was ranked first due to lowest AIC.Also, the proportion of the total predictive power found in the model was sorted to be the highest at weight value.The analysis was carried out for each grass parameter separately.

RESULTS
In total, the study identified 22 grass species and 10 species of sedges at Point Calimere (Table 2).The grass species dominated the stand in all parameters studied compared to sedges.Aeluropus lagopoides had the highest mean height, percentage of cover, soft texture, and green leaves.Chloris barbata had the highest percentage of reproductive phase.Dactyloctenium aegyptium was the second highest in terms of cover, soft texture green leaves, followed by Cyperus compressus.

Variation in grass parameters between season and among habitats
The grass parameters varied significantly among the three habitats: grassland, open-scrub, and dry evergreen.Among the five grass parameters, grass height, percentage cover, and reproductive phase were significantly more in open-scrub followed by grasslands and dry evergreen (Figure 2).On the other hand, softtextured grass and green leaves were significantly higher in grasslands than the other two habitats.In relation to season, grass height, percentage cover, soft texture, and green leaves were significantly higher in the wet compared to dry season.While the reproductive stage was significantly higher during dry compared to wet season (Table 3).

Influence of covariates on grass parameter -GLM Model
Grass height: GLM to test the influence of covariates on grass height showed a model with covariates viz feralhorse density, P. juliflora cover, rainfall, spotted deer density, shrub % cover, and distance from water bodies turned out as the best model with lower delta AIC and higher weightage, influencing the grass height (Table 4).However, the covariates, viz., feral-horse with negative influence and rainfall with positive influence were alone turned out to be the significant predictors of grass height (Table 5).
Grass cover: Although the GLM showed that covariates with feral-horse density, grasslands, P. juliflora cover, rainfall and distance from water bodies entered as the best model with lower delta AIC and higher weightage influencing the grass cover (Table 4), covariates, viz., feral-horse, P. juliflora cover, distance from water with negative influence, and rainfall and wet season with positive influence turned out to be the significant predictors of grass cover (Table 5).

Grass soft-texture:
A model with open habitat, P. juliflora density, rainfall and distance from shade entered as the best model with lower delta AIC and higher weightage influenced the soft-texture of grass (Table 4).Interestingly, all the four covariates had significant effect on the soft-texture with rainfall, open habitat, and distance from shade were having positive influence, while the P. juliflora density had negative influence (Table 5).
Grass green leaves: Although a model with blackbuck density, open habitat, P. juliflora density, and rainfall entered as the best model with lower delta AIC and higher weightage influencing the green leaves availability (Table 4), covariates, viz., density of P. juliflora and blackbuck with negative effect and open habitat with positive effect were alone significantly influencing the percentage of green leaves in the grass species (Table 5).
Grass reproductive phase: Although a model with covariates such as feral-horse density, open habitat extent, P. juliflora density, and rainfall entered as the 23194 J TT best model with lower delta AIC and higher weightage influencing the reproductive phase of the grass (Table 4), covariates, viz., P. juliflora density and open-scrub with positive effect and rainfall with negative effect had significant influence on the amount of grass reproductive phase (Table 5).

Grass dynamics parameters Description
Dependent variables 1 Grass height (cm) Grass height was measured using a measuring scale, from the ground level to the highest leaf blade bend, at five points (one each at four corners and one at the center) of the quadrate.

2
Grass cover/m 2 Assessed visually assuming 100% for the entire quadrat and estimating the proportion of area within a quadrat covered by each grass.
3 Soft texture (%) Examined crushing the leaves by hands, if leaf's structure could be squashed into a ball-proportion of such leaves for a given grass species in quadrat was rated in % rating.

4
Green leaves (%) Assessed visually quantifying the proportion of leaves in a given species with green grass, assuming 100% for all the leaves of the same species.

5
Reproductive phase (%) Evaluated visually quantifying the proportion of a given grass with flowers and fruits in % rating.

Covariates 6
Open habitat extent (km 2 ) At every plot laid, habitat visibility on all four directions of north, south, east, and west.

7
Distance to water (m) Measured as the distance from a given quadrate to the water source using a rangefinder or obtained from land-use land-cover map. 8 Distance to shade (m) Measured as the distance from a given quadrat to the nearest canopy cover area using a rangefinder.

9
Distance to road (m) Measured as the distance from a given quadrat location to the nearest road or obtained from land-use land-cover map.
10 Ambient temperature (°C) Measured using a generic digital thermometer-cum-hygrometer device (model: HT01) at each observation at the feeding site.
12 Weather Recorded visually as cloudy or sunny weather at the start of each feeding site examination.

Predictors of grass height
The density of feral horses, a non-ruminant bulk feeder (Arandhara et al. 2020), had the greatest negative effect on grass height among the 13 covariates compared.According to Maron & Crone (2006), the effects of herbivory on grassland are more severe than those on woodland.Rainfall, widely regarded as the most effective factor in promoting plant growth, was a major predictor influencing positively on height of the grass species (Derner & Hart 2007;Parton et al. 2012).Increased rainfall during the growing season has been shown to improve soil water use, which in turn promotes healthy root development and grasses as well as other plant growth (Wan et al. 2002).

Predictors of grass cover
The study showed that percentage of grass cover decreased with feral horse density, P. juliflora cover and distance from the waterbody.The feral-horse is a large herbivore with predominant grazing nature and a bulkfeeder (Baskaran et al. 2019;Arandhara et al. 2020).Their intensive grazing pressure is thus negatively influencing the grass cover, as grazing mostly occurs during the growing season (Hao & He 2019).In this study, a decrease in grass cover was observed with the Prosopis cover.Prosopis is an alien invasive species at Point Calimere, which grow taller and tap the sunlight at canopy level.Sunlight is an essential factor for the photosynthesis of all plants including grass species and thus the increase in Prosopis cover reduces the intensity of sunlight available to the grass species found at the ground level.Therefore, grass cover decreased with Prosopis cover (Baskaran et al. 2019;Murugan et al. 2019;Arandhara et al. 2021).Like the sunlight, soil nutrient, soil moisture is also another important factor influences the plant growth and productivity and moisture in the soil is required during the wet season to promote CO 2 absorption and plant growth (Morgan et al. 2016).Therefore, the grass  cover increased significantly with rainfall and during wet season compared to dry season as reported elsewhere (Wan et al. 2002;Zhang et al. 2020;Xu et al. 2021).Since soil moisture decreases with increase in distance from waterbody, the grass cover decreased significantly with distance from the waterbody.

Predictors of soft-texture and green grass
The study showed that grass soft-texture increased with rainfall, open habitat, and distance from the shade, but it decreased with Prosopis density.Similarly, the green grass availability increased with rainfall, and open habitat, but it decreased with Prosopis and blackbuck densities.Studies have shown that rainfall by increasing the soil moisture, triggering the growth of fresh, and green grasses (Hermance et al. 2015;Moore et al. 2015;Morgan et al. 2016;Post & Knapp 2020).The fact that fresh grown plants parts are softer than the old-grown parts due to low fiber and cellulose content ( de Jong 1995;Treydte et al. 2011;Kunwar et al. 2016).Thus, rainfall increases significantly both soft-texture and green leaves of grasses.The open habitat provides the ideal sunlight intensity and temperature for promoting photosynthesis in grass species (Solofondranohatra et al. 2018) and thus the soft-textured green grasses increased significantly with extent of open habitats.In contrast, with increase in Prosopis density, which exploits both sunlight at canopy level and the available soil moisture more efficiently (Shiferaw et al. 2021), reduces the softtexture and greenness of the grass species through reduced growth.The negative effect of shade on grass soft-texture also follows the above concept as explained for Prosopis.Blackbucks are a species of ruminant that is known for eating the tips of young, tender grass leaves, which are richer nutrients and water content (Jhala 1997;Baskaran et al. 2016) and thus green grass availability decreases with the density of blackbucks.

Predictors of grass reproductive phase
The study shows that the grass species reproductive phase increased with P. juliflora cover and openscrub.As stated in earlier studies that the open-scrub J TT predominantly with woody plants that are typically less than 3 m tall and relatively open, offers excellent support for grasses and plants that are shorter than them (Wardle 1971;Solofondranohatra et al. 2018).Likewise, P. juliflora is the most common woody plant in open-scrub at Point Calimere (Arandhara et al. 2021).Since, grazers are unable to access the grass species found between and beneath the bushes, the grazing intensity is lower in open-scrub compared to grasslands.Thus, the grass species with less grazing pressure in the open-scrub or with more density of Prosopis, with better growth as found in this study, had more reproductive phases than that of in grasslands.These findings are similar to earlier study that states the grass species in the open-scrub were shielded from overgrazing (Popay & Field 1996).As a result, the grass species reaches its maximum potential for growth and reproduction.As the mean annual precipitation across space increased, flowering time pushed back for most grass species, as documented by Munson & Long (2017).

CONCLUSIONS AND RECOMMENDATIONS
Point Calimere supports grasses and sedges which provide ideal food sources for mammalian grazing communities.Both grass species quantity and quality varied among habitats and between seasons.Among the 13 covariates compared, Prosopis, an alien invasive species, is the major driver that negatively influences on both grass quantity and quality.The feral horse, an alien invasive, negatively affected grass height.The devastating effect of these exotics on native flora and fauna at Point Calimere have been already documented by various studies (Ali 2005;Baskaran et al. 2019;Arandhara et al. 2020Arandhara et al. , 2021)).Thus, to safeguard the natural communities of plants and animals of Point Calimere, effective measures are needed as suggested by earlier studies.

Figure 1 .
Figure 1.Map showing the study area at Point Calimere Wildlife Sanctuary, India.

Figure 2 .
Figure 2. Variation in grass dynamics parameters among the habitats at Point Calimere Wildlife Sanctuary, India.

13
Rainfall (mm) Rainfall data arrived from the secondary sources (https://www.soda-pro.com/web-services/meteo-data/merra)14 Prosopis juliflora cover/25 m 2 Prosopis density were arrived from 5 x 5 m quadrates in the study area 15 Prosopis juliflora density/25 m 2 Prosopis cover were obtained from 5 x 5 m quadrates in the study area 16 Blackbuck density Density was obtained by the line transect survey method in the study area.22 species of grasses and 10 sedges in the study area, similar to what Arandhara et al. (2021) and Frank et al. (2021) reported for the same area.On the basis of cover, species such as A. lagopoides, D. aegyptium, and C. compressus are the three dominant grasses recorded in the coastal habitats, open-scrub, and dry-evergreen in this study.These findings also support earlier studies elsewhere for A. lagopoides (Khan & Gulzar 2003; Ahmed et al. 2013), D. aegyptium (Rojas-Sandoval 2016), and C. compressus (Ravi & Mohanan 2002; Bryson & Carter 2008).