W. Paul Sullivan | USGS Science

USGS2021

Trade-offs between suppression and eradication of sea lampreys from the Great Lake

Ecosystem managers confronted with newly invasive species may respond with a program of suppression or eradication. Suppression of an invasive species refers to management of a species such that its effect on other biota in the local ecosystem is acceptable. Eradication is the removal of all individuals of a species from a defined region. We examine the cost and benefit trade-offs between suppression and eradication of Laurentian Great Lakes sea lampreys ( Petromyzon marinus ) based on discussions at the 3rd Sea Lamprey International Symposium (held in 2019). Substantial effort has been expended annually since the 1960s to suppress sea lampreys in the Great Lakes basin. Choosing between suppression and eradication is a value judgement, ideally made jointly by scientists, decision-makers, stakeholders, and society. Successful large-scale eradications have been limited to a small number of cases for which the cost to human society justified and supported the long-term commitment necessary for success. The greatest challenge to successful eradication of sea lampreys from the Great Lakes may be a suitable social, political, legal, and institutional environment. Preparations could be made now for a transition in which public pushback on current control methods (pesticide applications and barriers to fish passage) leads to more extensive use of an alternative control method, such as genetic control.

Great Lakes and Saint Lawrence River areas↗

USGS2003

Relative contributions of sampling effort, measuring, and weighing to precision of larval sea lamprey biomass estimates

We developed two weight-length models from 231 populations of larval sea lampreys (Petromyzon marinus) collected from tributaries of the Great Lakes: Lake Ontario (21), Lake Erie (6), Lake Huron (67), Lake Michigan (76), and Lake Superior (61). Both models were mixed models, which used population as a random effect and additional environmental factors as fixed effects. We resampled weights and lengths 1,000 times from data collected in each of 14 other populations not used to develop the models, obtaining a weight and length distribution from reach resampling. To test model performance, we applied the two weight-length models to the resampled length distributions and calculated the predicted mean weights. We also calculated the observed mean weight for each resampling and for each of the original 14 data sets. When the average of predicted means was compared to means from the original data in each stream, inclusion of environmental factors did not consistently improve the performance of the weight-length model. We estimated the variance associated with measures of abundance and mean weight for each of the 14 selected populations and determined that a conservative estimate of the proportional contribution to variance associated with estimating abundance accounted for 32% to 95% of the variance (mean = 66%). Variability in the biomass estimate appears more affected by variability in estimating abundance than in converting length to weight. Hence, efforts to improve the precision of biomass estimates would be aided most by reducing the variability associated with estimating abundance.

Journal of Great Lakes Research↗