1. The range of a species is controlled by biotic and abiotic factors; both could have changed recently due to human activity.2. We used environmental modelling, morphometric and genetic data to interpret ecological responses at the species boundary of a pair of New Zealand grasshoppers with very different ranges; one widespread (Phaulacridium marginale) and one restricted to semi-arid central/southern South Island (Phaulacridium otagoense).3. Climate-and habitat-based distribution models for grasshoppers in the past (last glacial maximum), present and future (2070), in concert with modelling of vegetation patterns imply range and demographic expansion of P. marginale and stability of P. otagoense.4. mtDNA sequence revealed four main lineages with pronounced differences in genetic diversity and geographical range. The widespread lineage associated with P. marginale revealed a signature of range expansion but regionally restricted lineages were geographically structured at a fine scale. Within the narrow geographical range of P. otagoense, three mtDNA lineages resulted in high diversity, more typical of large stable populations.5. Geometric analysis of pronotum shape identified individuals from a region of sympatry with mixed characteristics. Mismatch of phenotype, mtDNA lineage and nuclear DNA sequence indicates introgression between grasshopper species now in contact. This appears to be accompanied by P. otagoense range reduction through ecological competition. 6. Deforestation by people starting $ 800 years ago best explains range change and resulting hybridisation of these grasshoppers. Anthropogenic habitat modification can have indirect consequences on insect biodiversity and conservation by enabling introgression between formerly separate populations and species.

Mitochondrial DNA sequence is frequently used to infer species' boundaries, as divergence is relatively rapid when populations are reproductively isolated. However, the shared history of a non-recombining gene naturally leads to correlation of pairwise differences, resulting in mtDNA clusters that might be mistaken for evidence of multiple species. There are four distinct processes that can explain high levels of mtDNA sequence difference within a single sample. Here, we examine one case in detail as an exemplar to distinguish among competing hypotheses. Within our sample of tree wētā (Hemideina crassidens; Orthoptera), we found multiple mtDNA haplotypes for a protein-coding region (cytb/ND1) that differed by a maximum of 7.9%. From sequencing the whole mitochondrial genome of two representative individuals, we found evidence of constraining selection. Heterozygotes were as common as expected under random mating at five nuclear loci. Morphological traits and nuclear markers did not resolve the mtDNA groupings of individuals. We concluded that the large differences found among our sample of mtDNA sequences were simply owing to a large population size over an extended period of time allowing an equilibrium between mutation and drift to retain a great deal of genetic diversity within a single species.