Time filter

Source Type

Haghi Vayghan A.,Gorgan University of Agricultural Sciences and Natural Resources | Zarkami R.,Guilan University | Sadeghi R.,Ghent University | Fazli H.,The Caspian Sea Ecology Research Center
Hydrobiologia | Year: 2015

Predicting and modeling of habitat preferences of fish is a very important issue for aquatic management. Classification trees (CTs) were used to predict the habitat preferences of the Caspian kutum (Rutilus frisii kutum, hereafter kutum) in the southern Caspian Sea. The applied model was optimized with genetic algorithm (GA) and greedy stepwise (GS) to select the most explanatory variables for predicting the presence/absence of kutum. The suitability index was considered to determine the quality and suitability of fish habitat in the sea. The results of Paired Student’s t tests showed that there was a significant difference between predictive performances of models before and after variable selection methods. Both optimizers improved the predictive power of CTs and resulted in a better understanding of CTs by making a selection of the sea characteristics that were used as inputs to the models. The results show that the effect of different seasons, sea depth, and photosyntheticaly active radiation were the main predictors affecting the habitat preferences of kutum in the Caspian Sea. Constructed trees in combination with GA and GS showed high capability when applied to predict the habitat preferences of this valuable commercial fish species. Determining the habitat needs of the target fish will enhance local fisheries performances and the long-term conservation planning of the fish to implement the ecosystem-based management in the Caspian Sea. © 2015 Springer International Publishing Switzerland

Haghi Vayghan A.,University of Tehran | Poorbagher H.,University of Tehran | Taheri Shahraiyni H.,Tarbiat Modares University | Fazli H.,The Caspian Sea Ecology Research Center | Nasrollahzadeh Saravi H.,The Caspian Sea Ecology Research Center
Aquatic Ecology | Year: 2013

The relationship between species and habitat is important in ecosystem-based fisheries management. Habitat suitability index (HSI) modeling is a valuable tool in ecology and can be used to describe the relationship between fish abundance and ecological variables in order to estimate the suitability of specific habitats. In the present study, an HSI model was applied to determine suitable habitats for the Caspian kutum (Rutilus frisii kutum), an important commercial species in the southern Caspian Sea. An arithmetic mean model (AMM) was found to be the most appropriate model for describing the relationship between two of the environmental variables investigated (depth and benthos biomass). However, a geometric mean model explained the evident relationship when all four environmental variables were used (depth, benthos biomass, photosynthetically active radiation and sea surface temperature). The areas with an HSI > 0.5 had over 85 % of the total catch indicating the reliability of the prediction of the Caspian kutum habitat using the AMM. The present study showed that depth and substrate structure are the most important environmental variables for the Caspian kutum to select its habitats, and between remotely sensed data, chlorophyll a, photosynthetically active radiation and sea surface temperature are the most critical parameters for near real-time prediction of the Caspian kutum habitat. © 2013 Springer Science+Business Media Dordrecht.

Discover hidden collaborations