Premium
RELATIONSHIPS OF STREAM INVERTEBRATE COMMUNITIES TO DEFORESTATION IN EASTERN MADAGASCAR
Author(s) -
Benstead Jonathan P.,
Douglas Michael M.,
Pringle Catherine M.
Publication year - 2003
Publication title -
ecological applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.864
H-Index - 213
eISSN - 1939-5582
pISSN - 1051-0761
DOI - 10.1890/02-5125
Subject(s) - streams , ecology , deforestation (computer science) , biodiversity , habitat , rainforest , geography , benthic zone , generalist and specialist species , national park , environmental science , biology , computer network , computer science , programming language
Madagascar has been recently identified as a global hotspot for freshwater biodiversity. Loss of most of its eastern rain forest, combined with a high incidence of micro‐endemism and specialization to forest stream habitats, has likely led to extinction of many of the island's stream insect species. We compared habitat and macroinvertebrate community structure in three streams draining protected rain forest within Ranomafana National Park in eastern Madagascar and three agriculture streams draining the park's largely deforested peripheral zone. Multivariate analyses showed that macroinvertebrate communities differed between stream types. Forest streams were characterized by species‐rich, diverse communities composed primarily of collector–gatherers and collector–filterers belonging to the orders Trichoptera, Ephemeroptera, Plecoptera, and Diptera. In contrast, relatively depauperate agriculture stream communities were dominated by generalist collector–gatherer taxa mostly belonging to the order Ephemeroptera. Agriculture streams had 0–14% canopy cover, significantly higher mean and maximum water temperatures, and significantly lower benthic organic matter storage than forest streams (71–83% canopy cover). We found no evidence for higher sedimentation rates or nutrient concentrations in agriculture streams. Differences in thermal regime may be important in the declines of certain taxa in agriculture streams (e.g., Plecoptera). However, our data support the hypothesis that differences in stream communities are caused mainly by differences among taxa in ability to track shifts in basal resources (i.e., terrestrial detritus vs. in situ algal production) caused by deforestation. Community simplification and loss of endemic specialist taxa will continue if projections of future deforestation are proved correct. We recommend that stream biodiversity be considered in future conservation planning and propose the creation and maintenance of multiuse riparian forests that could serve both ecological and economic functions. Given that more than 5 × 10 5 km of stream and river channel are affected by tropical deforestation annually, our results have potential applications in other tropical regions.