How does stream degradation impact invasive species?

Methods

Summary

Seven Great Lakes tributaries across Michigan were selected as representative stream habitats for round goby (Au Sable, Clinton, Muskegon, Ocqueoc, Rifle, Rouge, and Stony Creek). They span a gradient of size, relative water quality, population density, and land use so as to represent the variation inhabitat associated with Round Goby invasion in Michigan. In each watershed, three sites were chosen which represent areas where Round Goby do and do not occur.

Sampling began in 2015 and will continue through 2017. Each site is visited once during the summer, as close to the fish reproductive season as possible. Each site is sampled for 1 to 1.5 hours until all available habitat is explored. At each site, the fish community is sampled with a 3x1.25m nylon mesh seine (3.18mm mesh). Captured fish are identified and released, except a subset of individuals (three per species as vouchers) which are brought back to the laboratory. 

For stable isotope analysis, tissues are taken from voucher individuals from the right caudal peduncle (the area of muscle just in front of the tail). This tissue is cleaned of any scales, skin, and bones, and then freeze dried. Each muscle sample is then ground to a fine powder using a mortar and pestle and contained in a tin capsule. These capsules are then sent to another facility where they will be analyzed for stable isotope composition (carbon and nitrogen).

Stable isotope analysis is done using a mass spectrometer. A mass spectrometer ionizes a sample before sending it through a magnetic field. As the sample goes through the field, the sample will be deflected by the magnet (think of it turning a corner). Particles which have the same mass will be deflected to the same degree. When they reach a detector on the other end, their placement indicates how much they were deflected, and thus the mass of the particle. The total number of particles to hit each place on the detector indicates the relative composition of each (heavy or light) isotope.

Challenges

Our hypothesis suggests there will be a difference in isotopic composition of Round Goby and their native competitors. We suggest this will be the case because we expect Round Goby to be a superior competitor in these rivers, as their increasing abundance would suggest. However, it is possible that we will not be able to find statistical difference between Round Goby and native species. Should this happen, it would indicate a generalist feeding strategy among all species assessed. Although this is contrary to our hypothesis, it is still informative. Overlap of dietary niches (as indicated by stable isotopes) would suggest that food resources are not limiting in these systems. If there is ample prey for both Round Goby and native competitors, then any interactions between species must exist in a different capacity. This would allow us to rule out competition for food resources as an impact of Round Goby on native species such that we could focus on other forms of interaction going forward (i.e., reproductive effort, habitat availability, etc).

Pre Analysis Plan

Many analytical approaches exist for stable isotope data. We will use an analytical approach described in Turner, Collyer & Krabbenhoft (2010). This method has been previously used by the author and can be easily run in the statistical software R. The procedure used will analyze the spatial layout of isotope data in a plot (carbon on the x-axis, nitrogen on the y-axis). Using a permutation procedure, it calculates the distance between the center point for each species, the overall spread of those points (how much of the plot each species takes up), and how well 'clustered' those points are. This will give us an overall idea of how much 'overlap' there is between species, and thus an indication of competition for diet items.