The effects of land use at different spatial scales on instream features in agricultural streams

• Autor(a) principal: Molina, Mariana C. [UNESP]
• Data de Publicação: 2017
• Outros Autores: Roa-Fuentes, Camilo A. [UNESP], Zeni, Jaquelini O. [UNESP], Casatti, Lilian [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1016/j.limno.2017.06.001; http://hdl.handle.net/11449/174795
• topic: Agriculture and Stream homogenization; Native forest; Pasture; Sugarcane

The conversion of forests into agriculture has been identified as a key process for stream homogenization. However, the effects of this conversion can be scale-dependent. In this context, our aim was to identify the influence of different land uses at different spatial scales (catchment, drainage network and local) on instream features in agricultural streams. We defined six classes of land use: native forest, reforestation, herbaceous and shrubs, pasture, sugarcane and other categories. We obtained 22 variables related to instream, riparian area, stream morphology and water physicochemical characteristics in 86 stream reaches. To identify and isolate the effect of different land uses at different spatial scales on instream features, we performed a partial redundancy analysis (p-RDA). Different land uses and scales influenced instream features and defined two stream groups: (i) homogeneous streams with a higher proportion of sand substrate and instream grasses that were associated with the proportion of herbaceous vegetation at the local scale and with pasture at all scales and (ii) heterogeneous streams with a higher physical habitat integrity associated with the proportion of forest and sugarcane at the local and catchment scales. Land use at the catchment scale affected the physicochemical water properties and stream morphology, whereas stream physical habitat (i.e., substrate, instream cover, marginal vegetation and stream physical habitat condition) was mainly influenced by land use at the local scale (i.e., 150 m radius). Pure catchment, drainage network and local land uses explained 9%, 7% and 4%, respectively, of the total variation of instream features. Thus, to be most effective, stream conservation and restoration efforts should not be limited to only one scale.