|dc.description.abstract||Healthy lake ecosystems support biodiversity and human populations. They provide many ecosystem services such as water, food and energy. Lakes can be impacted by natural disturbances, but they are increasingly threatened by human-induced disturbances. Studies have shown that eutrophication and climate change often enhance cyanobacteria over other photosynthetic taxa. As cyanobacterial blooms are becoming more frequent and intense throughout the world, more lake systems are being investigated. In some cases there is not a clear link between eutrophication and cyanobacterial blooms. One such example is Lake Pounui (Wairarapa, New Zealand), which has little intensive agriculture in its catchment but water quality has degraded markedly in the last decade. The lake now experiences heavy cyanobacterial blooms every summer. This could be due to the presence of a non-native fish population, the European perch (Perca fluviatilis). This thesis examined the relationship between cyanobacterial blooms and perch introduction in New Zealand
lakes, including a multi-trophic study in Lake Pounui. Perch were introduced c. 1870 in New Zealand but introduction records are patchy and sometimes non-existent. Moreover, most lake systems are not studied until they are already degraded. This thesis used a combination of traditional proxies (pollen, charcoal, pigments) and modern proxies (sedimentary ancient DNA, XRF scanning) from lake sediment cores to reconstruct lake ecology in pre-human times, after M¯aori settlement between the 13th to 15th century, and after European settlement from 1840 AD. Timelines and intensity of human impact were reconstructed with pollen, charcoal analysis, and sediment dating when possible.
Cyanobacterial communities in six lakes were reconstructed through their sedimentary ancient DNA (sedaDNA) using metabarcoding and droplet digital PCR (ddPCR) in Chapter 2. Bloom-forming species were present in all lakes prior to human arrival; however their overall abundance was low. Total cyanobacteria abundance and richness increased in all lakes after European settlement but was very pronounced in four lakes, where bloom-forming taxa became dominant. The trends in cyanobacterial abundance from ddPCR were then compared to cyanobacterial pigments (canthaxanthin, echinenone, myxoxanthophyll and zeaxanthin) using highperformance liquid chromatography in Chapter 3, to assess the likelihood of the historical increase observed. Pigments / sedaDNA relationships were more consistent when all pigments were summed, which is likely due to differences in species composition across lakes. The positive correlations confirmed an increase in cyanobacterial biomass since European arrival.
Due to patchy records for fish introduction, fish sedimentary DNA was compared to environmental DNA (eDNA) from water samples as a methodological check (Chapter 4) before applying this method to the sediment cores. This study was undertaken in three small and shallow lowland lakes by targeting perch and rudd (Scardinius erythrophthalmus). Fish DNA was evenly distributed across the whole lake except when the fish population was low. Samples collected from the sediment contained fish DNA more often than water samples in two out of the three small shallow lakes (including Lake Pounui). Sediment geochemistry probably impeded detection in the third lake. Perch sedaDNA was therefore used as an indication of fish presence in Lake Pounui for Chapter 5, which explored multitrophic changes in Lake Pounui over the last c. 1,000 years. In addition to pollen, charcoal, and 14C dating, XRF scanning was used to reconstruct mineralogic shifts from the catchment (Ti/inc, K/inc) and within the lake (inc/coh). Biological trends were reconstructed
by targeting the sedaDNA of bacteria (16S rRNA), microeukaryotes (18S rRNA), metazoans (CO1), and macrophytes (rbcL, trnL). Complemented by historical records and studies, the data produced in this thesis indicated that the biggest changes in Lake Pounui happened after European settlement (c. 1845), with land clearance, perch introduction, climate change, and probable fertiliser application driving the degradation of the water quality in c. 180 years. This study revealed shifts in native communities (macrophytes, bacteria, oligochaete worms) and the appearance of new species (perch, macrophytes, freshwater nematodes) previously undocumented using sedaDNA. The results highlight just how complex yet fragile lake ecosystems can be and how little we still know about them. Sedimentary ancient DNA is a useful tool to study the insidious and long-lasting impact of nonnative species on freshwater ecosystems because it widens the range of species that can be studied. However, it needs to be complemented with other proxies. This thesis provides a framework to study fish DNA in small shallow lakes (Chapter 4). It can also inform future management and restoration strategies in lakes, especially in Lake Pounui, by retracing historical water quality (Chapter 2) and identifying taxa present prior to, during, and after lake degradation (Chapter 5).||