USGS2015
We examined how the frequency of attacks by Sea Lamprey on fishes in Lake Ontario varied in response to Sea Lamprey abundance and preferred host abundance (Lake Trout >432mm). For this analysis we assembled seven data sets. Two fishery independent surveys for Lake Trout: US Geological Survey (USGS)/New York State Department of Environmental Conservation (NYSDEC) south shore September gillnet assessment of adult Lake Trout (USGS/NYSDEC SGNS)( Lantry and Lantry 2011); and Ontario Ministry of Natural Resources (OMNR) monthly June-November community index gillnetting in northeastern Lake Ontario (OMNR CIS) (Ontario Ministry of Natural Resources 2011). One angler survey: NYSDEC April-September Fishing Boat Survey data collected along the south shore for Chinook and Coho salmon, and Rainbow and Brown trout (NYSDEC FBS) ( Lantry and Eckert 2012). Two spawning run datasets: OMNR north shore data including spring spawning runs of Rainbow Trout in the Ganaraska River and electroshocking data for fall spawning runs of Chinook and Coho salmon in the Credit River (Ontario Ministry of Natural Resources 2011); and NYSDEC data from the Salmon River on the southeastern shore including October spawning runs of Chinook and Coho salmon. One Sea Lamprey spawning survey: Department of Fisheries and Oceans Canada (DFO)/US Fish and Wildlife Service (USFWS) data for spawning-phase Sea Lamprey abundance collected from known spawning streams distributed throughout the Lake Ontario drainage basin (Mullet et al. 2003). One assessment of the abundance of dead Lake Trout: USGS/NYSDEC October-November bottom trawl collection of Lake Trout carcasses (Schneider et al. 1996). Annual patterns in A1, A2 and A3 wound stages did not track well in plots of wounding rates for the USGS/NYSDEC SGNS and correlations between A1 and later stages did not exist. A1 rates were not correlated to either Lake Trout abundance or Sea Lamprey numbers when considered alone, but were strongly correlated to the ratio between Sea Lamprey numbers and Lake Trout abundance (parasite/host ratio). While A2 and A3 rates were correlated to each other, neither was consistently correlated to any of the Lake Trout abundance or Sea Lamprey abundance parameters and sums of A1 to A3 rates did not improve correlations over those for A1 rates considered alone. Our analysis of the strain-specific susceptibility of Lake Trout to attack by Sea Lampreys extended the previous Schneider et al. (1996) analysis of three strains (SUP, CWL, and SEN) and 11 years of data 1982-1992 to an analysis of seven strains (SUP, CWL, SEN, JEN, LEW, ONT, and OXS) and two groups of unmarked fish (1983-1995 and 1996-2010) and included 18 more years of data through 2010. The susceptibility to attack for CWLs and SENs were below SUPs and nearly identical to the earlier values, new values for LEWs were greater than SUPS and values of unmarked Lake Trout prior to 1996 were unexpectedly greater than SUPs. By reexamining the Schneider et al. (1996) regression relationship between A1 wounding on Lake Trout and the incidence of Lake Trout carcasses recovered in fall bottom trawls (including three additional years of data), and substituting A1 wounding rate for total numbers of A1s observed which was used as the independent variable in the previous version, we were able to increase the variance explained by the relationship from an r2 of 0.60 to 0.88. Healing rate of wounds was explored by examining the monthly incidence of A1 and A2 wounds on Lake Trout from the OMNR CIS. Because wounding intensity varied between years and monthly sample size was frequently low, the ratio of A2 to A1s wounds was used to index how wounds accumulated or disappeared from the Lake Trout populations across seasons. The A2/A1 ratio decreased between June/July and October. A simple wounding model for Lake Trout was constructed to simulate the pattern of ratios by inflicting wounds on the population each month according to a predetermined distribution and including parameters for lethality (fixed distribution) and healing rate of A1 and A2 wounds (discretely varied between simulations). The best simulated representation (ratio size and monthly pattern) of the OMNR CIS data came from an A1 healing rate of 0.5 months and an A2 rate of 2 months. To examine whether alternate hosts provided reliable data to index damage caused by Sea Lampreys we compared September values of abundance and A1 wounding rates on Lake Trout, Sea Lamprey abundance, and the parasite/host ratio with NYSDEC Creel data for observations of Sea lampreys attached to sport-caught salmonids and to wounding observations for salmonids sampled in OMNR and NYSDEC spawning run assessments. Attachment frequency on NYSDEC Creel Chinook Salmon, Brown Trout, and Rainbow Trout were strongly correlated to all measures of Lake Trout abundance and wounding and to the parasite/host ratio. Chinook Salmon and Coho Salmon wounding observations for the NYSDEC Salmon River spawning run assessments and Rainbow Trout from the OMNR Ganaraska River spawning run assessment were strongly correlated to wounding measures for nearly all salmonids and to the parasite/host ratio.