Abstract
Ecosystems have rapidly changed as humans have overtaken natural areas and replaced them with urbanisation. As a result of human settlement and habitat destruction, Bird-Window Collisions (BWCs) have become the second-most common cause of avian deaths. One of the primary reasons birds collide with windows is because they mistake the reflection of vegetation in the window for a natural corridor. As they attempt to fly through the corridor, they collide with a window. Many drivers of BWCs, such as geospatial, temporal, sex class, age class, seasonal, monthly, and decade trends, have not been researched thoroughly, which makes finding a solution to this problem difficult. One of the difficulties with determining whether species are experiencing BWCs more in one sex than the other is that 60% of birds are monomorphic, including the kererū (Hemiphaga novaeseelandiae).
The kererū is endemic to New Zealand and considered a taonga species. This large pigeon species is one of two bird species in New Zealand that disperse seeds larger than 12mm large seeds and can often be found in urban areas feeding on native and exotic plants. The kererū is a monomorphic species, and many methods for sexing this bird have been attempted; however, the accuracy of some methods remain unclear. Therefore, this study aimed to determine the most reliable field-based way to sex kererū and understand geospatial, temporal, sex class, age class, seasonal, monthly, and decade trends amongst kererū BWCs in Dunedin. The information gained from this study can increase our ability to manage and protect urban kererū throughout New Zealand and improve our knowledge of threats to kererū as urbanisation continues to expand.
The Otago Museum, Komiti Taoka Tuku Iho, and the Wildlife Hospital, Dunedin provided 100 dead kererū for this study. These 100 individuals were sexed by morphometric measurements, post-mortem examination of the gonads, and DNA sexing. An additional five live kererū were also measured and DNA sexed, as was one feather sample, which was obtained from a kererū after colliding with a residential property. DNA sexing was able to identify sex for all 106 samples and was assumed to be the most reliable method of sexing in this study. Post-mortem gonadal examination methods could not identify 22% of individuals due to the inability to locate gonads and further misidentified 16.7% of the sample that did have visible gonads. Morphometric measurements were able to determine sex with 80% accuracy by using a discriminant function analysis. These results indicate that DNA sexing is the most reliable method for sexing kererū.
This study also evaluated several geospatial, temporal, sex class, age class, seasonal, monthly, and decade trends in kererū deaths and injuries in Dunedin using a dataset collected since 1968. Using ArcGISPro and R Studio, current and emerging hot spots for kererū BWCs and specific spatial aspects likely to cause more collisions than others were determined. This study did not detect temporal, sex, or age class trends amongst kererū collisions in Dunedin. However, there was a significant correlation between kererū BWCs and areas containing dense residential properties and commercial buildings. Lastly, this study discusses methods to mitigate kererū BWCs in current and emerging hot spots.
This thesis has provided an accurate means for sexing kererū and has identified geospatial trends within kererū BWCs in an urban environment. With the knowledge that residential and commercial areas are significant hot spots for kererū BWCs, wildlife managers can use this information to continue monitoring kererū collision trends within New Zealand cities and determine how and where to mitigate current and future kererū collisions. DNA sexing provides a quick and reliable method to sex kererū and has the potential to be of use if their numbers decline and a conservation program for kererū is created. Data used throughout this study should continue to be collected to repeat this study in the future as urbanisation continues to expand in New Zealand. Finally, knowledge gained via the continuation of data collection and repeating this study in the future could promote more precise results for determining geospatial, temporal, sex class, age class, seasonal, monthly, and decade trends amongst kererū BWCs and can identify updated information on emerging hot spots in Dunedin.