Mangrove habitat suitability modeling: implications for multi-species plantation in an arid estuarine environment.

In southern Iran, Sirik Estuary hosts the only two-species (Rhizophora mucronata and Avicennia marina) mangrove forest in the northwesternmost edge of the Indian Ocean mangrove distribution. Aiming to protect its forest reserve and compensate for inevitable losses, this study utilized habitat suitability modeling (the Maxent model) to identify suitable afforestation zones for each species, independently. The model was calibrated using the location of successfully established mangrove saplings as presence points and an array of physical and sediment physio-chemical layers as predictive variables. The model yielded an acceptable training AUC value of 0.963 for A.marina and 0.982 for R.mucronata. Moreover, physical variables had the highest contribution to predicting suitable habitats with different levels of importance for each species. The majority of A.marina suitable habitats were distributed along the in-estuary creek banks, creating mangrove-lined waterways while the R.mucronata suitable habitats were mostly distributed at the base of the main water creeks in the seaward reaches of the estuary. According to the Mann-Whitney U test results, there was a statistically significant spatial niche segregation (z = - 12.14, p = 0.000, sig ≤ .05, 2-tailed) between the species' suitable habitats. The results showed that white mangroves tend to create mangrove-line structures along the water creeks penetrating inside the estuary while red mangroves mostly prefer the seaward side of the existing mangrove patches which are in danger of sea level rise.