For the majority of the last 20 Ma Banks Peninsula existed as an isolated island from the rest of New Zealand. As a result, numerous endemic species have evolved in this isolated region. Regional endemism has been recorded in plant, insect, and bird species. Unfortunately, a number of these species may now be extinct. Prior to human arrival the Peninsula was almost completely covered in dense mixed podocarp forest. However, human deforestation reduced indigenous forest on the Peninsula to < 1 % of its original coverage. Now native forest and scrub cover an estimated 15 % of Banks Peninsula (much of this is regenerated). Currently, Banks Peninsula has 10 regionally endemic stream invertebrates (nine formally described), which are all classified as ‘threatened’ or ‘at risk’. However, the conservation status of many New Zealand stream invertebrates is limited by a lack of scientific knowledge. Prior to this study it was unknown whether Banks Peninsula’s regionally endemic stream invertebrates were restricted to a few catchments and highly threatened or if they were just poorly studied. Previous work had indicated that several of the Peninsula’s endemic species were restricted to forested headwaters and forest fragments.

The primary aim of this study was to determine the distribution of Banks Peninsula’s regionally endemic stream invertebrates. Secondly, my study aimed to test whether regional and river classifications could explain the spatial distribution of these endemic species. Lastly, I aimed to assess stream invertebrate diversity across three spatial scales and determine whether selected taxa (including endemic species) showed specific microhabitat preferences on the Peninsula. I carried out a survey of 54 streams spatially separated across Banks Peninsula. At each stream benthic invertebrates were collected, and physico-chemical parameters were measured. To my knowledge this is the largest scale survey that has been carried out specifically assessing Banks Peninsula’s regionally endemic stream invertebrates.

I collected seven of the 10 known endemic species. Two endemic species, the caddisfly Hydrobiosis styx and the stonefly Zelandobius wardi were found in all three of the Peninsula’s Ecological Districts. However, the other five species I collected, including the caddisfly Costachorema peninsulae, the mayfly Nesameletus vulcanus, the net-winged midge Neocurupira chiltoni, the beetle Orchymontia banksiana, and the undescribed stonefly Zelandoperla sp.1 (BJF00160: Banks Peninsula) were restricted to Akaroa and Herbert Ecological Districts. Two species (O. banksiana and N. chiltoni) were common and widely distributed across the Peninsula. However, the remaining five endemics occurred in isolated, scattered, and small sub-populations, which probably reflect the areas historic deforestation.

The endemic species tended to occur in two main community groups. While N. chiltoni, O. banksiana, and C. peninsulae were characteristic of communities that occurred in larger, non-forested streams. H. styx, N. vulcanus, Z. wardi, and Zelandoperla sp. 1 preferred smaller higher altitude streams with higher amounts of native riparian vegetation and shading. Further, analysis using the River Environment Classification (REC) suggested that these seven endemic species occur commonly in cold wet environments within steep small to medium sized streams sourced from hills. In contrast the third classification system I tested, the Freshwater Ecosystems of New Zealand (FWENZ), was not able to robustly explain the occurrences of the majority of these endemic species.

I also investigated diversity across the Peninsula at three scales – alpha (local or stream diversity), beta (catchment) and gamma (regional). I collected a total of 95 taxa. This high diversity is similar to the stream invertebrate diversity of other much larger ecoregions in New Zealand. Alpha diversity showed a significant pattern of increasing from the north and west towards the south and east. Local stream diversity was most strongly associated with larger streams (wider, faster velocities, and deeper). However, the absence of clear patterns between beta diversity and measured physico-chemical parameters suggests that larger scale environment heterogeneity is influencing beta diversity. Endemic species accounted for 7 % of the Peninsula’s (gamma) diversity and up to 21 % the diversity in individual streams (alpha diversity). Furthermore, several of the regionally endemic species showed strong preferences for specific microhabitats, particularly riffles, suggesting these endemics are highly adapted to the Peninsula’s steep gradient streams.

Lastly, I review the conservation status of seven regionally endemic stream invertebrates. Based on my assessment, the status of N. vulcanus and Z. wardi is better than currently listed and their classification could be changed to ‘Nationally Vulnerable’. The status of four endemic species (C. peninsulae, H. styx, N. chiltoni, and O. banksiana) seemed appropriate. However, my results indicate that Zelandoperla sp. 1 should have a higher threat status and be reclassified as ‘Nationally Endangered’. Although forested headwater streams on the Peninsula do not have as diverse habitats as larger lowland streams, they contributed markedly to the diversity of Banks Peninsula and are important habitats for the regionally endemic species. Therefore, the protection of forested headwaters is critical to preserve the diversity of Banks Peninsula and it’s regionally endemic stream invertebrates.