We selected 60 northern temperate lake sites in Vilas County, Wisconsin lake district. Methods for lake choice and sampling are given in greater detail in Marburg et al. (2005) Each lake was sampled once between 2001 and 2004, in June, July, or August (15 different lakes each summer). We chose stratified lakes deeper than 4 m to insure that all the lakes contained a diverse fish community. With two exceptions (chains of lakes), lakes were chosen to be in separate watersheds. Lakes were chosen based on two criteria: landscape position, using historical DNR water conductivity data as a proxy of position, and riparian housing development, measured in buildings km-1 shoreline (Marburg et al. 2005). Landscape position refers to the location of a lake along the hydrological gradient. The gradient ranges from the top of a drainage system, where seepage lakes are fed mainly by rainwater, through lakes which receive water from groundwater and have surface outflows, to lakes further down in the drainage system, which receive water from both ground and surface flow (Kratz et al. 1997).
Landscape position affects lake water chemistry, because as water flows across the surface and through soil, it picks up carbonates and other ions which increase the water’s electrical conductivity (specific conductance, a temperature-independent measure of salinity), alkalinity, and its ability to support algal and macrophyte production. In addition, aspects of lake morphology correlate with landscape position. Most obviously, larger lakes tend to occur lower in drainage systems (Riera et al. 2000).
The riparian (near-shore terrestrial) zone around northern Wisconsin lakes is being rapidly developed for use as both summer and permanent housing (Peterson et al., 2003). Concurrent with housing development, humans often directly and indirectly remove logs (Kratz et al. 2002) and aquatic vegetation (Radomski and Goeman 2001) from the littoral zone (near shore shallow water area), resulting in reduced littoral zone complexity. The slowly-decaying logs of fallen trees create physical structure (coarse woody habitat: CWH) in the littoral zone of lakes that provides habitat and refuge for aquatic organisms (Christensen et al. 1996). Fish, including plankton-eating species (planktivores), reproduce and develop in shallow water (Becker 1983). Because planktivorous fish affect zooplankton community structure through size-selective predation (Brooks and Dodson 1965), there is the potential for indirect effects of housing development on zooplankton.
Lakes ranged in size from 24 to 654 ha. In 2001, 2002 and 2004 we chose lakes from the extreme ends of the conductivity and housing density gradients and in 2003 lakes were chosen to fill in the gap in the middle of the ranges. The study lakes range from oligotrophic to mesotrophic (Kratz et al. 1997; Magnuson et al. 2005).
At each lake we sampled zooplankton, water chemistry, riparian and littoral vegetation, fish, crayfish, and macrophytes. Each lake was sampled only once, but given the large number of lakes sampled in this area, we expect to see relationships between variables within lakes and at a landscape scale. A snapshot sampling design maximizes sites that can be visited, and is sufficient for a general characterization of zooplankton communities (Stemberger et al. >001).
Environmental Sampling and Analysis:
Physical, chemical and biological samples were taken above the deepest point in each lake during the summer stratification period (June, July, or August). Water samples were collected from one half meter depth using a peristaltic pump, and were analyzed for pH, alkalinity, specific conductance, water color, chlorophyll-a, dissolved organic and inorganic carbon, total phosphorus, and total nitrogen (Appendix Table 1). Secchi depth, temperature and dissolved oxygen profiles, and vertical plankton tows were also taken at the deepest point. Temperature and dissolved oxygen concentrations (DO) were measured through the water column at 1 meter increments.. Conductivity, TP/TN, alkalinity and pH water samples were collected unfiltered while water for DIC/DOC and color water samples was filtered through nucleopore polycarbonate filters. Alkalinity, pH, and DIC/DOC samples were filled to the top and sealed quickly to prevent CO2 loss or invasion. Samples containing air bubbles were recollected. Chlorophyll samples were collected on glass fiber filters in the field. Water chemistry and chlorophyll a analyses were done at the Trout Lake Biological Station, Boulder Junction, WI except for TP, TN, DIC and DOC samples, which were analyzed at the Center for Limnology-Lake Mendota Laboratory, Madison, WI.
Zooplankton Sampling and Identification:
Each zooplankton sample was taken at approximately the deepest part of each lake, via a vertical tow with a Wisconsin net (30” long, 80µ mesh) lowered to 1.0 meter above the bottom of a lake and then pulled up slowly at a rate of about 3 seconds per meter. Contents of the net were washed into 4-oz jars with 95% ethanol. A replicate sample was taken from each lake.
Zooplankton samples were analyzed for planktonic crustacean and insect species. In most cases, approximately 5000 individual zooplankters were scanned and identified to species. For some lakes, the entire sample was scanned, because fewer than 5000 individuals were present. Samples were scanned using a dissecting microscope under high and low magnifications. Adult representatives of each species found per subsample were dissected and mounted on slides using Hoyers mounting medium (Dodson and Frey, 2001). Cladocerans and copepods were identified to species using resources cited in Schell et al. (2001). Presence/absence of each species was used in the ordination analyses.
Littoral Zone Surveys:
Littoral habitat, fish and macrophyte surveys were performed at eight sites within each of the 55 lakes. The sites were chosen by randomly selecting two points per compass quadrant of each lake. Each year littoral habitat surveys were conducted in June, fish surveys in July and macrophyte surveys in August.
Littoral habitat (substrate and coarse woody habitat) was measured along a 50 m transect parallel to shore along the ½ meter depth contour at each site. The two Littoral CWH variables (number of logs km-1 > 5 cm diameter, and number > 10 cm) were transformed by log of (1+number) to normalize the variables.
Littoral fish were sampled in July of each year, along the shallow areas (water depth > 0 and <2 m) adjacent to the riparian plots. Night electroshocking and crayfish and minnow traps were used to catch fish and crayfish. All species were identified and counted.
At each site, we examined the littoral vegetation along a transect perpendicular to the shoreline. Within a ¼ m2 quadrat at every meter mark, we recorded the total percent vegetation cover, dominant species and all species present to a water depth of 2 meters or 50 meters from shore. If a depth of 2 m was reached before 20 quadrats were measured, a second transect was performed 25 m to the right of the initial point. These observations were averaged to calculate % total cover of vegetation in the littoral zone per lake (Marburg et al. 2005). The % cover was transformed using arcsine square root of the decimal proportion.