Premium
Can changes in ant diversity along elevational gradients in tropical and subtropical A ustralian rainforests be used to detect a signal of past lowland biotic attrition?
Author(s) -
Burwell Chris J.,
Nakamura Akihiro
Publication year - 2016
Publication title -
austral ecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.688
H-Index - 87
eISSN - 1442-9993
pISSN - 1442-9985
DOI - 10.1111/aec.12322
Subject(s) - rainforest , transect , species richness , ecology , biological dispersal , subtropics , range (aeronautics) , tropical rainforest , biodiversity , species diversity , biology , geography , population , materials science , demography , sociology , composite material
Increasing temperatures are predicted to have profound effects on montane ecosystems. In tropical forests, biotic attrition may reduce lowland diversity if losses of species due to upslope range shifts are not matched by influxes of warmer‐adapted species, either because there are none or their dispersal is impeded. Australian rainforests consist of a north–south chain of patches, broken by dry corridors that are barriers to the dispersal of rainforest species. These rainforests have repeatedly contracted and expanded during Q uaternary glacial cycles. Many lowland rainforests are expansions since the L ast G lacial M aximum and may, therefore, show a signal of historical biotic attrition. We surveyed ants from replicated sites along three rainforest elevational transects in eastern A ustralia spanning 200 to 1200 m a.s.l. and nearly 14° of latitude. We examined elevational patterns of ant diversity and if there was possible evidence of lowland biotic attrition. Each transect was in a different biogeographic region; the A ustralian W et T ropics (16.3° S ), the central Q ueensland coast (21.1° S ) and subtropical south‐eastern Q ueensland (28.1° S ). We calculated ant species density (mean species per site) and species richness (estimated number of species by incorporating site‐to‐site species turnover) within elevational bands. Ant species density showed no signal of lowland attrition and was high at low and mid‐elevations and declined only at high elevations at all transects. Similarly, estimated species richness showed no evidence of lowland attrition in the W et T ropics and subtropical south‐east Q ueensland; species richness peaked at low elevations and declined monotonically with increasing elevation. Persistence of lowland rainforest refugia in the W et T ropics during the L ast G lacial M aximum and latitudinal range shifts of ants in subtropical rainforests during the Holocene climatic optimum may have counteracted lowland biotic attrition. In central Q ueensland, however, estimated richness was similar in the lowlands and mid‐elevations, and few ant species were indicative of lower elevations. This may reflect historical biotic attrition due perhaps to a lack of lowland glacial refugia and the isolation of this region by a dry forest barrier to the north.