International Lake Ontario-St. Lawrence River Study

About Us

News/Media

Newsletter

Public Interest Advisory Group

Technical Work Groups

Reports and Minutes

Study Data

Links

The Boardroom

International Joint Commission

Great Lakes Information Network
Web Site and Translation
by the Translation Bureau





Get Adobe Reader
Download Adobe Reader 7.0
Technical Working Groups

Northern Pike - young-of-year (YOY) net productivity (Upper St. Lawrence River - Thousand Islands area)

Performance Indicator Summary


PI Name/Short Description: Northern Pike - young-of-year (YOY) net productivity (Upper St. Lawrence River - Thousand Islands area) [E17]

Technical Workgroup: Environmental TWG

Research by: John M. Farrell, Jerry V. Mead, and Brent Murry

Modeled by: Jerry V. Mead, John M. Farrell, Brent Murry, LTI (DePinto, Redder)

Performance Indicator metrics: Relative YOY production (g/ha/year) is simulated for three habitats (drowned river mouths, and non-protected and protected bays, and shoals) from spring spawning to the late summer YOY period. A young-of-year (YOY) Northern pike production model is used as a performance indicator to assess effects of water level and temperature variation on critical early life stages. The growth and abundance (production) of YOY are known to be the primary forces in determining the strength of year classes that maintain northern pike populations. The performance indicator is a spatially explicit individual based model that integrates life stage specific (egg, fry, juvenile) growth, relative abundance, survival and production for three primary habitats. Spawning is simulated over the entire littoral gradient, spanning from seasonally flooded emergent vegetation in wetlands (drowned river mouth) and bays (open and protected) to permanently flooded submersed aquatic vegetation in bays and shoals (~0.2 to 6 meters or ~0.66 to 19.69 ft water depth).

Ecological Importance/Niche: Northern pike are the dominant piscivore in littoral habitats for the upper St. Lawrence River and are a top predatory species with a strong influence on fish communities system wide. Northern pike influence both size and population dynamics of yellow perch, which dominate fish biomass, and many other fish species they consume. Northern pike are an excellent indicator species due to their role in the fish community, and their dependence and sensitivity to wetland habitats during spawning and early life history. Due to their sensitivity, Northern pike are a good indicator of system changes and their populations have experienced significant population declines.

Temporal validity The PI is representative of early life history processes occurring from early spring (March 1st) through summer (August 23rd). The model endpoint matches long-term monitoring data for YOY Northern pike that is used for model calibration and validation. Model predictions will be most accurate for the Post-Seaway era conditions for which data for model development were available.

Spatial validity The Northern pike YOY performance indicator is valid for the upper St. Lawrence River Thousand Islands Region. The PI was developed from field data collected in 16 study areas within this region. Bathymetric digital elevation models (DEM) were created specifically for these study sites with high resolution. Water depths from site specific gauging locations were used to test predictions of depth from DEMs.

Hydrology Link: Research indicates that springtime water levels that enhance northern pike spawning success were historically important, but today appear to be decoupled from age-0 production and subsequent year-class formation (Farrell 2001). This may be due to current water level management practices preventing access of spawners to preferred habitat types and potentially stranding eggs following spawning. A secondary effect of hydrologic management is long-term habitat changes, including the increase of cattail (Beland 2003; Farrell et al. 2003, Halpern et al. 2003), and the loss of sedge meadow habitats have likely influenced northern pike reproductive success. Post-Seaway year class strength models indicate greater importance for late summer/fall water levels, where low levels promote stronger year-classes, rate of spring warming (days until 8C is reached), and summer temperatures (#days>20C or 68F). These factors are consistent with a post-Seaway habitat change and access scenario for spawning northern pike, and suggest deeper, later spawning and a stronger role of nursery habitat conditions.

Algorithm: Northern pike eggs are deposited using as a function of mean daily water temperature and vegetation coverage (grasses, sedge meadow, and submerged aquatic vegetation) with a Logistic regression probability model. Vegetation coverage in wetlands is predicted using the ETWG wetland submodel; for elevations less than 74.2 m (243.44 ft) (IGLD 1985) submerged aquatic vegetation was mapped by interpreting high resolution ortho-images of study sites taken in the spring of 2003 (NYS GIS Clearinghouse), and low elevation springtime aerial photographs (SUNY College of Environmental Science and Forestry). Loss of eggs and yolk larvae by stranding due to water level fluctuation is incorporated in model function. Rate of development and survival of eggs to swim up of larvae are simulated daily as a function of habitat specific water temperature and time (days) determined from laboratory trials (Farrell and Toner 2002). Growth of swim-up larvae, following complete yolk absorption and start of exogenous feeding, is simulated using a temperature and consumption driven bioenergetics model. A habitat specific proportion of maximum consumption achieved was fit with field estimates of northern pike growth. Habitat specific mean daily survival of YOY northern pike is predicted as a function of body length and was determined from field studies and literature values. Production for each habitat is the simulated product of growth and abundance for August 23.

Calibration Data: Calibration of the model is extensive including field egg density data collected over the three habitat types, egg development and mortality trials conducted to predict timing of swim-up of larvae, and a YOY bioenergetics model developed specifically for the period of first piscivory to emigration.

Validation Data: The northern pike spawning component of the model is validated by comparison of predicted observations to field data collected from independent historical data sampling for egg density. For egg deposition probability relationships to vegetation density and water temperature, 15% of the raw data is held out of model construction and used in validation. Predicted and observed growth and abundance data are compared for replicate study areas and temporal periods. Year class strength regression relationships to model predictions of YOY production are also used in model validation. A sensitivity analysis is performed on simulated northern pike production relationships with water temperature to assess the relative influence of major model components.

Documentation and References: The primary source of data collected in the field for development of this performance indicator originates from long term datasets on Northern pike from the SUNY College of Environmental Science and Forestry (ESF). ESF maintains field data on Northern pike in a partnership with NYS Department of Environmental Conservation. Important and unique datasets including field collections of naturally spawned eggs (Farrell 2001), monitoring of abundance of YOY northern pike, and completion of 12 related thesis and dissertation research projects. In addition, two recent studies completed for the Great Lakes Protection Fund involving coastal wetland habitats (Farrell et al. 2003a) and the use of spawning marshes (Farrell et. al.2003b) provides important data regarding northern pike and their critical habitats.

  • Beland, M. 2003. Holocene vegetation dynamics of an upper St. Lawrence River coastal wetland and surrounding uplands: Effects of climate change and anthropogenic disturbance. SUNY College of Environmental Science and Forestry, Master's Thesis, 167 pages.

  • Farrell, J. M. 2001. Reproductive success of sympatric northern pike and muskellunge in an Upper St. Lawrence River bay. Transactions of the American Fisheries Society 130:796-808.

  • Farrell, J. M., D. J. Leopold, A. D. Halpern, J. A. Toner, B. A. Murry, and M. Beland, 2003. Restoration of coastal wetlands in the St. Lawrence River through re-establishment of natural hydrologic regimes. SUNY College of Environmental Science and Forestry, Syracuse, NY. Final Report Submitted to the Great Lakes Protection Fund, Evanston, Ill., 79 pages.

  • Farrell, J. M. and J. A. Toner. 2003. Fish Recruitment: Evaluation of Hydrologic Management Effects on Northern Pike and Muskellunge Performance in Lake Ontario and the Upper St. Lawrence River. SUNY College of Environmental Science and Forestry, Syracuse, NY. Year 2 Report to the International Joint Commission. 15 pages.

  • Farrell, J. M. and A. D. Bosworth. 2003. Use of Spawning Marshes in Rehabilitation of St. Lawrence River Northern Pike and Muskellunge Populations. SUNY College of Environmental Science and Forestry, Syracuse, NY. Final Report submitted to the NYS Great Lakes Protection Fund, Buffalo, NY. 36 pages.

  • Halpern, A.D., J. M. Farrell, Jason A. Toner, Molly Beland, Brent A. Murry and Donald J. Leopold 2003. Can water-level control structures restore function and diversity of wetlands? (New York) Ecological Restoration 21(4): 317-318.

Risk and uncertainty assessment: The major assumptions of the YOY production indicator are listed below:

  • The 16 study areas in the upper river are representative of the entire upper St. Lawrence River.

  • Egg and yolk-fry have 100% mortality if habitat suitability becomes zero.

  • Variability in egg fertilization was assumed equal among habitats and over years.

  • Spawning habitat area polygons in submersed vegetation types are of constant area and coverage in elevations less than 74.2 m (243.44 ft) (IGLD 1985).

  • P value (the proportion of maximum consumption achieved) was a fixed, but estimated value for each habitat and does not exhibit interannual variation in prey availability.

  • Precision of spatial simulations of water temperature (1d) will limit model performance.

  • Conditions for which the model was developed are applicable to the temporal period being simulated.


Top of page