A project undertaken at The University of Queensland, and supervised by David Merritt.
The endemic glow-worms of Australia and New Zealand comprise nine species. They are the larvae of a type of fungus gnat in the genus Arachnocampa. The bioluminescent larvae tend to be found in colonies and are the subject of tourism at several locations. Glow-worm populations are often highly fragmented because they are confined to wet caves, moist forest gullies and the vicinity of waterfalls. Their generally poor ability to fly could mean that gene flow between locations is restricted, so we hypothesised that populations would be highly structured genetically.
To test this hypothesis, we collected larvae from multiple sites, including caves, gullies and waterfalls for three species of Arachnocampa; one Australian mainland species, Arachnocampa girraweenensis, the Tasmanian endemic, Arachnocampa tasmaniensis, and the New Zealand endemic species, Arachnocampa luminosa. We sequenced nuclear DNA (nDNA) using RADseq and estimated connectivity and diversity. We also analysed mitochondrial DNA (mDNA) based on sequencing of the COI gene, known as the barcoding gene. mDNA is maternally inherited so it gives an indication of female dispersal. In New Zealand, the project team collected specimens from 28 sites covering both main islands.
Preliminary analysis indicates that populations are highly structured, with no strong evidence of gene flow over large distances. The COI mDNA shows high diversity and high site specificity. There is no sharing of mDNA haplotypes between populations, except between individuals of a cluster of four populations located 80 km or less apart in Fiordland in south-western Otago. Overall, the COI data are consistent with there being little to no successful dispersal over large distances. Further analysis is required before we can assess whether nuclear gene flow is higher than the female-restricted mitochondrial gene flow.
We suggest that haplotype sharing in the Fiordland cluster is due to a past founder effect since the area was subject to Pleistocene glaciation and has become suitable habitat only in the last 10,000 years. Thus, recolonisation could have been recent, explaining the shared haplotypes. Similar “founder effects” explanations for COI haplotype sharing have been suggested in other insects of the region.
A second feature of the New Zealand sampling is the stark genetic divergence of a cluster of three populations in the south-east Otago region of the South Island. They form a divergent cluster in both mtDNA and nDNA analyses and may represent a distinct species. Investigations looking for species-specific morphological features are being carried out by Julia Kasper at Museum of New Zealand, Te Papa Tongarewa.
The team collected specimens of Arachnocampa tasmaniensis, the Tasmanian glow-worm, from 18 locations. The RADseq data analysis and the completed COI dataset indicates two or three geographic clusters. Once again, there is little haplotype sharing between sites: most locations within the identified geographic clusters have unique but closely related haplotypes. Individuals taken from different caves over a limited geographic range and even from inside and outside a cave are separable with principal component analysis, indicating that populations within caves show low dispersal. Forest populations tend to show higher levels of admixture. Together, the data indicate a pattern consistent with isolation-by-distance and limited female dispersal.
The Australian mainland species, Arachnocampa girraweenensis, is found in northern New South Wales and southern Queensland in rainforest sites and in caves. This species also shows a high degree of structure between populations even those from a cluster of granite boulder caves that are less than 20 km apart. In summary, glow-worms of the three species show similar high diversity and differentiation of mtDNA haplotypes with most sites having unique haplotypes. A few sites have shared haplotypes, but this is rare.
We conclude that each examined Arachnocampa species is genetically diverse, indicating that populations are localised with infrequent gene flow between sites. This is interpreted as being due to the low mobility of glow-worms, especially females, which are weak fliers and appear to restrict their egg-laying to the location where they themselves emerged as adults.