Freshwater ecosystems face unprecedented levels of human-induced stresses and it is expected that the invasion of non-indigenous species will cause the greatest loss of biodiversity in lakes and rivers worldwide. Bythotrephes longimanus is a predatory invertebrate that invaded North America in the early 1980s, first being detected in the Great Lakes, and then moving to a number of inland lakes in Ontario and the northern United States. Using experimental and survey-based approaches, I tested several hypotheses concerning the effects of Bythotrephes on native zooplankton community structure and function. My results indicate that Bythotrephes reduces total abundance, biomass, and richness of zooplankton, especially cladoceran taxa, throughout the ice-free season. As a result of high predation pressure by the invader, total seasonal and epilimnetic zooplankton production was also substantially reduced in invaded lakes, which may have important consequences for the transfer of energy to fish and other taxa that feed on zooplankton. Interestingly, there was some evidence that zooplankton shifted their reproduction in time and space to avoid Bythotrephes, which may buffer the effects of the invader on food web functioning. Other measures of ecosystem function were relatively unaffected by the invasion of Bythotrephes. In addition, Bythotrephes may interact in unexpected ways with other anthropogenic stressors, and act to slow down the process of recovery by preying on species that maintain community abundance during acidification, but also affecting species attempting to recolonize historically acidified lakes. Although dispersal of zooplankton may maintain some of the ecosystem functions provided by zooplankton communities, loss of biodiversity may be a permanent result of invasion. The effects of the continued spread of invasive species across the landscape may be profound, as the invader Bythotrephes has demonstrably altered zooplankton communities and may reduce the ability of freshwater ecosystems to respond to future environmental change and maintain ecosystem functioning. / Thesis (Ph.D, Biology) -- Queen's University, 2007-07-19 14:56:57.102

Bythotrephes longimanus is an invasive, macroinvertebrate from Eurasia that was introduced into the Great Lakes region in the mid 1980s. Bythotrephes introductions into lake ecosystems have resulted in substantial changes in zooplankton communities, including declines in species richness, abundance, biomass and production. Changes in zooplankton communities may alter the quantity and quality of prey to other predators such as cisco (Coregonus artedii), a pelagic forage fish. Here, I conduct a current day comparison of cisco populations to determine if prey consumption by cisco differs in the presence of Bythotrephes, and whether changes in diet result in energetic consequences (changes in growth and condition) for cisco. Effects of Bythotrephes on native zooplankton communities have resulted in substantial changes in the variety and proportion biomass of zooplankton and macroinvertebrate prey types in cisco stomachs, which have in turn modified growth of cisco. Cisco taken from invaded lakes achieve greater total lengths but changes in condition were not detected. This effect may be driven by improved growth in the second and subsequent growing seasons, suggesting that growth consequences for young fish (that do not feed on Bythotrephes) are different than for older individuals. Length-at-structure age data indicate that by the end of the first growing season (age 1) cisco achieve comparable total body lengths in invaded and reference lakes, suggesting that food consumption by young cisco remains unchanged by Bythotrephes. Alternatively, young cisco forage may be reduced in the presence of Bythotrephes, resulting in decreased survival and similar growth among individuals that survive to age 1. In contrast, despite changes in the zooplankton community; growth of older fish (≥ age 2) was enhanced in lakes that have Bythotrephes. Improved growth among older cisco (≥ age 2) in invaded lakes may be related to the presence of a newly attainable, high energy prey source (Bythotrephes). Alternatively, enhanced growth may be explained by lower competition due to reduced recruitment of young cisco (≤ age 1) in invaded lakes. Increased knowledge regarding the effects of Bythotrephes on growth of cisco is important in furthering our understanding of its impact on lake ecosystems. / Thesis (Master, Biology) -- Queen's University, 2010-04-28 22:46:07.756

Human-mediated introductions of non-indigenous species now threaten to homogenize the biota of the Globe, causing huge economic and ecological damage. This thesis studies the ecological role of 3 invasive planktonic crustaceans, the omnivorous copepod Acartia tonsa (western Atlantic and Indo-Pacific) and the predatory cladocerans, Cercopagis pengoi (Ponto-Caspian) and Bythotrephes longimanus (Eurasian). B. longimanus invaded the North American Great Lakes in 1982, C. pengoi the Baltic in 1992 and the Great Lakes in 1999, while A. tonsa has an extensive invasion history that includes the Baltic. We review current knowledge on feeding biology of the predatory cladocerans. A study of stable C and N isotope ratios indicated mesozooplankton as the main food source of C. pengoi in the northern Baltic Sea proper, with young C. pengoi also eating microzooplankton, such as rotifers. Young-of-the-year herring did eat C. pengoi and herring trophic position shifted from 2.6 before the invasion to 3.4 after, indicating that C. pengoi had been “sandwiched” into the modified food web between mesozooplankton and fish. Salinity tolerance experiments on Acartia tonsa and co-occurring Acartia clausi showed the formers euryhaline character and high grazing potential. Energy partitioning between ingestion, production and respiration was rather constant over the tested salinity range of 2 to 33, with small differences in gross growth efficiency and cost of growth, but maximum ingestion at 10-20. Egg hatching in A. tonsa was only reduced at the lowest salinity. Extreme changes in salinity were needed to cause significant mortality of A. tonsa in the field, but its feeding activity could be severely reduced by salinity changes likely to occur in estuaries. A study of a hypertrophic estuary showed that A. tonsa can sustain a population despite very high mortality rates, caused by predation, high pH and low oxygen, helping explain the success of A. tonsa as an invader of estuaries.

When we develop and use species distribution models to predict species' current or potential distributions, we are faced with the trade-offs between model generality, precision, and realism. It is important to know how to improve and validate model generality while maintaining good model precision and realism. However, it is difficult for ecologists to evaluate species distribution models using field-sampled data alone because the true species response function to environmental or ecological factors is unknown. Species distribution models should be able to approximate the true characteristics and distributions of species if ecologists want to use them as reliable tools. Simulated data provide the advantage of being able to know the true species-environment relationships and control the causal factors of interest to obtain insights into the effects of these factors on model performance. I used a case study on Bythotrephes longimanus distributions from several hundred Ontario lakes and a simulation study to explore the effects on model performance caused by several factors: the choice of predictor variables, the model evaluation methods, the quantity and quality of the data used for developing models, and the strengths and weaknesses of different species distribution models. Linear discriminant analysis, multiple logistic regression, random forests, and artificial neural networks were compared in both studies. Results based on field data sampled from lakes indicated that the predictive performance of the four models was more variable when developed on abiotic (physical and chemical) conditions alone, whereas the generality of these models improved when including biotic (relevant species) information. When using simulated data, although the overall performance of random forests and artificial neural networks was better than linear discriminant analysis and multiple logistic regression, linear discriminant analysis and multiple logistic regression had relatively good and stable model sensitivity at different sample size and detection limit levels, which may be useful for predicting species presences when data are limited. Random forests performed consistently well at different sample size levels, but was more sensitive to high detection limit. The performance of artificial neural networks was affected by both sample size and detection limit, and it was more sensitive to small sample size.

When we develop and use species distribution models to predict species' current or potential distributions, we are faced with the trade-offs between model generality, precision, and realism. It is important to know how to improve and validate model generality while maintaining good model precision and realism. However, it is difficult for ecologists to evaluate species distribution models using field-sampled data alone because the true species response function to environmental or ecological factors is unknown. Species distribution models should be able to approximate the true characteristics and distributions of species if ecologists want to use them as reliable tools. Simulated data provide the advantage of being able to know the true species-environment relationships and control the causal factors of interest to obtain insights into the effects of these factors on model performance. I used a case study on Bythotrephes longimanus distributions from several hundred Ontario lakes and a simulation study to explore the effects on model performance caused by several factors: the choice of predictor variables, the model evaluation methods, the quantity and quality of the data used for developing models, and the strengths and weaknesses of different species distribution models. Linear discriminant analysis, multiple logistic regression, random forests, and artificial neural networks were compared in both studies. Results based on field data sampled from lakes indicated that the predictive performance of the four models was more variable when developed on abiotic (physical and chemical) conditions alone, whereas the generality of these models improved when including biotic (relevant species) information. When using simulated data, although the overall performance of random forests and artificial neural networks was better than linear discriminant analysis and multiple logistic regression, linear discriminant analysis and multiple logistic regression had relatively good and stable model sensitivity at different sample size and detection limit levels, which may be useful for predicting species presences when data are limited. Random forests performed consistently well at different sample size levels, but was more sensitive to high detection limit. The performance of artificial neural networks was affected by both sample size and detection limit, and it was more sensitive to small sample size.