Generalized additive models to predict adult and young brown trout (Salmo trutta Linnaeus, 1758) densities in Mediterranean rivers
Habitat suitability models (HSM) are concerned with the abundance or distribution of species as a consequence of interactions with the physical environment. Generalized Additive Models (GAMs) were used to model brown trout (Salmo trutta L.) density as a function of environmental variables at the sca...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Fecha de publicación: | 2016 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | inglés |
| OAI Identifier: | oai:riunet.upv.es:10251/64821 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/64821 |
| Access Level: | acceso abierto |
| Palabra clave: | Generalized additive models Salmo trutta Habitat suitability models Hydromorphological units Mesohabitat TECNOLOGIA DEL MEDIO AMBIENTE |
| Sumario: | Habitat suitability models (HSM) are concerned with the abundance or distribution of species as a consequence of interactions with the physical environment. Generalized Additive Models (GAMs) were used to model brown trout (Salmo trutta L.) density as a function of environmental variables at the scale of river reach and hydromorphological units (HMU) in the Jucar River Basin (Eastern Spain). After 4years of observations (2003-2006) the data representing trout density were split into two categories, young (<2years) and adult (2years), for modelling independently. The environmental descriptors at reach-scale described the geographical position, hydrological conditions, proportions and diversity of habitats. At the scale of HMUs (pool, glide, riffle or rapid), habitat descriptors representing dimensions, substrate, cover and velocity were used. The best and parsimonious GAM for each category was selected after a comprehensive trial of all possible combinations of input variables. The models explained 61% (adult) and 75% (young) of the variability of the data (R(2)adj). The results demonstrated the relevance of mean width, mean depth, cover index, mean velocity and slope for adult brown trout. Young trout densities were mainly related to maximum depths, cover index, mean velocity, elevation, average distance between rapids and number of slow water HMUs. This article shows the relevance of considering geographical and habitat-related requirements at different scales to describe the patterns of trout density. Furthermore, the importance of considering non-linear relationships with habitat variables was demonstrated. The results are useful for environmental managers to design effective and science-based restoration measures, and result in a more efficient management of brown trout populations. |
|---|