Species richness of birds along a complete rain forest elevational gradient in the tropics: Habitat complexity and food resources matter

AIM: We examined whether the available surface area, temperature, or habitat complexity (foliage height diversity index) determine species richness of birds (and species richness of individual feeding guilds) along a complete forest elevational gradient. Further, we focused on the relationship between species richness of insectivorous birds and the availability of their food resources. LOCATION: Elevational gradient (200–3,700 m) of Mt Wilhelm (4,509 m a.s.l.), Central Range, Papua New Guinea. TAXON: Birds. METHODS: We collected data on bird communities at eight sites (500 m elevational increment) during three surveys encompassing both dry and wet seasons over a 2‐year period. We used point counts, mist‐netting, and random walks throughout a standardized area. We tested three predictors of diversity and all of their combinations, in conjunction with sensitivity analyses for spatial effects. Habitat complexity (foliage height diversity index) and temperature were locally measured; surface area available within 200 m elevational intervals was obtained using GIS software. We further locally surveyed insect biomass and related it to species richness of insectivorous birds. RESULTS: Birds displayed a monotonic decline in species richness (from 113 to 35 bird species) with increasing elevation, and a nested pattern of species loss. The observed patterns were best explained by habitat complexity for the insectivores, frugivore‐insectivores, and total number of bird species. The available surface area was the best predictor for frugivorous birds. The mean temperature had a high correlation with species richness of all birds and gave the best fit of species richness for insectivore‐nectarivores and pure nectarivores. The biomass of insectivorous birds correlated with the biomass of arthropods. We ruled out the possibility that the elevational pattern observed in birds could be driven by a single phylogenetic radiation. MAIN CONCLUSIONS: We observed species richness patterns correlate well with habitat complexity and mean temperature, but mean temperature was not ranked as high as expected. Our results thus challenge the generally expected high importance of temperature as a regulator of water availability, production, and biochemical process that influence species richness, and underscore the importance of vegetation structure and the food resources as the driver of observed species richness.