Biogeography of butterflies in the Australian monsoon tropics

Abstract

The biogeography of butterflies within the monsoon tropical biome of northern Australia is reviewed in terms of patterns of species richness, endemism and area relationships. Available data indicate that the region supports a relatively rich fauna, comprising 265 species (∼62% of the total Australian fauna), but endemism is low (6%). No genera are endemic to the monsoon tropics, but two (Neohesperilla, Nesolycaena) are characteristic components, embracing a total of seven species in the region, of which five are endemic. Three ecological specialists (Neohesperilla senta, Elodina walkeri, Candalides delospila), each associated with different vegetation types, appear to be characteristic elements of the monsoon tropics. Of 67 range-restricted species in the monsoon tropics, 15 (mostly associated with savanna) are endemic to the region, while 52 (mostly associated with rainforest) are non-endemic, occurring also in south-east Asia and/or mainland New Guinea. A pronounced attenuation in species richness from Cape York Peninsula across the Top End to the Kimberley is evident. Within the monsoon tropics, Cape York Peninsula stands out as an area of exceptional biodiversity, with 95% of the butterflies (251 species; 7 endemic species, 31 endemic subspecies/geographical forms) recorded from the entire region, compared with the Top End (123 species; 3 endemic species, 17 endemic subspecies/geographical forms). In contrast, the Kimberley has a comparatively depauperate fauna (85 species; 1 endemic species, 0 endemic subspecies) without strong Indonesian affinities, and contains only two range-restricted species. A sister-area relationship between Cape York Peninsula and the Top End-Kimberley is evident in one clade, Acrodipsas hirtipes (northern Cape York Peninsula) + A. decima (Top End), with a pairwise divergence of ∼1% based on mtDNA, and is suspected in another, Nesolycaena medicea (southern Cape York Peninsula) and N. urumelia (Top End) + N. caesia (Kimberley); a further five species show similar sister-area relationships across the Carpentarian Gap but at the level of subspecies or geographical form. Three general and complementary hypotheses are proposed to explain patterns of geographical differentiation of butterflies in the monsoon tropics: (1) the Carpentarian Gap is a biogeographical filter, functioning as a barrier for some species but as a bridge for others; (2) divergence among taxa between Cape York Peninsula and the Top End-Kimberley has occurred fairly recently (Quaternary), probably through vicariance; and (3) the Bonaparte Gap, with the exception of Nesolycaena, is not a vicariant barrier for butterflies in the Top End and Kimberley.