3.3. Influence of Environment and Seagrass Structure on Epifaunal Abundance
All seagrass variables including epiphytic algal biomass were significant in explaining variation in epifaunal abundance (PERMANOVA: DistLM, p < 0.01, Table 3.4). Initially, distance-based linear models were run without including leaf width due to its strong correlation with leaf length. However, a better fit was achieved when leaf width was included. Among the seagrass variables, leaf width contributed the highest proportion of variation in epifaunal abundance, followed by leaf length, density, biomass, and epiphyte biomass (Table 3.4). All seagrass variables were selected in the first model that best explained variability in epifaunal abundance determined by lowest AIC and highest R2 values. The second best model which included four variables (density, leaf width, epiphyte and seagrass biomasses), and had an AIC value < 2 levels from the first model, also adequately explained the observed variation (Table 3.4).
Environmental variables assessed independently in a DistLM marginal test, found exposure to contribute the highest proportion of variation followed by chl a and turbidity (Table 3.5). Temperature, salinity, and oxygen accounted for smaller proportions of variation (Table 3.5). All environmental variable contributions were significant (p < 0.001). The top two models based on AIC and R2 parameters, that best explained variation in epifaunal abundance selected five (all) and four (excluding turbidity) environmental variables respectively (Table 3.5).
The final structural equation model (SEM) employed to explore the influence of environmental variables and seagrass metrics on epifaunal abundance revealed several noteworthy findings. The direct effects of temperature, salinity, pH, and exposure on abundance were found to be non-significant (chi-square = 90.66, df = 45, p = 0.35, RMSEA = 0.083; Fig. 3.4B). However, turbidity exhibited a strong negative direct effect (-0.85), while oxygen displayed a positive direct effect (1.09). Indirectly, temperature, pH, exposure, and oxygen had negative effects on epifaunal abundance, whereas turbidity had a positive indirect effect (Table 3.6). Although all environmental variables directly influenced leaf length, this did not translate into indirect effects on abundance. In contrast, leaf width was negatively influenced by temperature and turbidity, and positively influenced by pH and exposure. It exerted a positive direct effect, explaining 65% of the variation in abundance (Fig. 3.4B, Table 3.6). Similarly, shoot density had a positive direct effect on abundance, accounting for 65% of variation (Fig 3.4B). Shoot densities were negatively influenced by pH, oxygen and exposure and positively affected by turbidity (Table 3.6). No other seagrass metrics emerged to directly influence epifaunal abundances.