A comparative analysis of univariate and multivariate spatial requirements modeling for habitat selection in freshwater fish: A case study of Oxynoemacheilus bergianus in the Jajroud River, Iran and implications for conservation and management

被引:0
|
作者
Mahmoudi, Saleh [1 ]
Abdoli, Asghar [1 ]
机构
[1] Shahid Beheshti Univ, GC SBU, Environm Sci Res Inst, Dept Biodivers & Ecosyst Management, Tehran, Iran
关键词
Habitat use; Habitat suitability index; Spatial modeling; Oxynoemacheilus bergianus; Jajroud River; TROUT SALMO-TRUTTA; CASPIAN SEA BASIN; SUITABILITY MODELS; ASSEMBLAGE STRUCTURE; KORDAN RIVER; SAVA RIVER; NEMACHEILIDAE; PREFERENCES; VELOCITY; IMPACT;
D O I
10.1016/j.ecolmodel.2025.111076
中图分类号
Q14 [生态学(生物生态学)];
学科分类号
071012 ; 0713 ;
摘要
Understanding the habitat requirements for fish in fluvial ecosystems is a key factor in conservation and making decisions for management. One of the main questions is what are the optimal ranges for each variable-velocity, depth, and substrate structure-that the species requires, and how do these ranges influence its development and survival? Given the scarcity of biological and ecological information on Oxynoemacheilus bergianus habitat selection at the mesoscale, this survey was designed to provide an answer to the aforementioned question. For this purpose, the effects of three main variables, including flow velocity, water depth, and substrate composition on the habitat selection of the species in a lotic ecosystem in northern Iran were evaluated. The results indicate that for the juvenile group, depths up to 20 cm, velocities of 15-20 cm.s-1, and a dominant substrate of fine and medium gravel are the most suitable ranges. For the adult group, depths of 20-30 cm, velocities of 15-30 cm.s-1, and small cobbles are preferred We addressed these questions using both univariate and multivariate approaches, which resulted in different variable importance, model accuracy, and uncertainty across methods. Specifically, flow velocity was identified as the most significant variable in the univariate analysis. The depth variable was more important in the multivariate approach. The minimum method (MI) and the arithmetic mean (AM) method had the lowest and highest error and uncertainty among the combined methods, respectively. The Generalized Linear Models (GLM) and Random Forest Models (RF) approaches showed the most accurate models in the Species Distribution Models (SDMs) for juvenile and adult groups, respectively. Utilizing these vital results will empower managers to make informed decisions aimed at conserving O. bergianus, ultimately leading to effective strategies that protect its habitat and ensure the species' long-term survival.
引用
收藏
页数:9
相关论文
empty
未找到相关数据