Multiple-stressor effects of agricultural stressors on stream benthic communities: Using genetic tools and biological traits to complement ecological research

Abstract

Most ecosystems are subject to multiple anthropogenic stressors acting simultaneously, impacting biodiversity across all levels from genotypes to ecosystems. Stressors may interact, resulting in non-additive effects that cannot be predicted from the effects of the individual stressors and can exacerbate the degradation of ecosystems and biodiversity loss. One of the greatest threats to freshwater biodiversity is agricultural intensification, with streams draining agricultural catchments impacted by elevated levels of nutrients, agricultural chemicals such as nitrification inhibitors, deposition of fine sediment on the stream bed and reduction in discharge and water velocity due to water abstraction for irrigation. For practical purposes, community measures and higher-level taxonomic groups are often the focus of studies assessing the impacts of stressors on stream ecosystems. This means key components of biodiversity are often overlooked. The use of genetic methods such as DNA barcoding and high-throughput sequencing may help overcome problems with morphological identification and allow the study of entire communities in their environment. Additionally, the use of species traits can provide insights into the mechanisms driving the effects on ecosystem functioning. To investigate the effects of stressors on stream ecosystems, I performed two surveys of 43 streams sites in southern New Zealand spanning wide gradients of agricultural stressors (nutrients and fine sediment) in autumn and spring. General linear models and an information-theoretic model selection approach were used to examine stressor relationships with benthic algal taxa and traits (Chapter 3), and species of the important invertebrate bioindicator taxa Potamopyrgus and Deleatidium (identified using DNA barcoding) (Chapter 2). These surveys were followed by two field experiments in 128 stream-fed outdoor mesocosms. The first investigated the individual and combined effects of the nitrification inhibitor DCD, along with nutrient enrichment and fine sediment on benthic algal taxa and traits (Chapter 4). The second experiment focused on mimicking dynamics in real streams where DCD concentrations peaked after rainfall events. Again, interactions of DCD with other stressors were examined, in this case nutrients, sediment and stream flow velocity. The individual and combined effects of these stressors on algal taxa and traits (Chapter 5) and benthic bacterial communities using high-throughput amplicon sequencing (Chapter 6) were investigated. The results of the stream surveys indicated that nutrients and sediment were influential stressors for the algal community and the two invertebrate bioindicator taxa. Algae also showed strong seasonal changes, with community composition changing significantly between the two sampling periods. Due to these seasonal differences and high variability of community composition between sites, algal traits tended to be a more useful tool than taxonomic composition for the investigation of stressor effects. With regard to the invertebrates, Potamopyrgus (consisting of a single species) showed positive relationships to nutrients and sediment, whereas Deleatidium abundance was unrelated to stressor levels when assessed at the genus level. However, Deleatidium was found to consist of 12 distinct genetically identified clades, with the three most abundant clades showing contrasting relationships to the stressors. These results indicate that species within a genus can differ substantially in their tolerance to stressors and respond in more complex ways than observed at the genus level, highlighting the benefits of including molecular techniques in ecological research. In the first experiment, sediment addition and nutrient enrichment (and their interactions) had pervasive effects on the algal community as expected. DCD, however, had relatively few and weak effects, despite experimental DCD concentrations including extreme levels above those found in real streams. In experiment two, which tested four stressors, sediment and flow velocity reduction had highly pervasive effects on the algal community. Nutrient addition was less pervasive, but effects were greater in magnitude and frequency than DCD. DCD addition applied at realistic concentrations and in realistic pulses, had fewer negative effects compared to the constant application and extreme concentrations tested in experiment 1. However algal communities in both experiments were affected by interactions between DCD and the other stressors, indicating that DCD could have negative effects on algal communities in streams impacted by multiple stressors. In contrast to the algal communities, the bacterial community in experiment 2 was strongly influenced by DCD. DCD is intended to target ammonium-oxidising bacteria in soil while having little effect on the overall bacterial community. For stream bacteria, however, DCD had the most pervasive effects on bacterial community evenness and prevalence of almost all common taxa. This was followed by sediment, then velocity reduction, whereas nutrient enrichment had surprisingly little effect. These results highlight the potential of threats by DCD to freshwater ecosystems. Stressor interactions, including synergisms and antagonisms, were found, indicating that interactions between these agricultural stressors need to be accounted for when assessing agricultural impacts on stream ecosystems.