METHODS

Study Area

NOCA is located along the crest of the Cascade Range of northern Washington, USA. There is no road access to NOCA lakes; they can be reached only by hiking or helicopter. Many lakes in NOCA were stocked with trout (primarily cutthroat, Oncorhynchus clarki, and rainbow, O. mykiss) during this century for recreational angling (Jarvis 1987).

Abiotic and Biotic Samples

Twenty-seven lakes were sampled two or more times in at least one year from 1989 through 1993 (Table 4.1). All zooplankton analyses were conducted using only these 27 lakes except determination of the effect of fish on large diaptomids in deep lakes (maximum depth >10 m), where four deep lakes that were sampled only once per year were also included to increase sample size. Lakes were sampled between ice-out (mid-June to early July) and the onset of inclement weather in the fall (approximately mid-September).

Table 4.1. Occurence of adult diaptomid copepods in North Cascades National Park Service Complex, 1989-1993.

For each lake on each sample date, three replicate vertical tows were taken with a 20-cm-diameter number 25 (64 µm mesh) zooplankton net. In 1989 only one vertical tow was taken in each lake. Vertical tows were taken near the deepest part of the lake. The net was lowered to within 1 m of the lake bottom and towed upward at a rate of about 0.5 m sec^-1. In the field, samples were preserved in 5% neutral sugar formalin (Haney and Hall 1973). In the laboratory, samples were split using a Folsom plankton splitter. A split portion was poured into a settling chamber and left to settle for 24 h. Organisms were identified to species. Organisms of each species were counted using an inverted microscope at 100X magnification. Body length (metasome and urosome) of adult diaptomids of each taxon was measured.

All analyses were conducted using densities of adult diaptomids. In Kettling and Juanita lakes, adult stages of copepods were not present when the lakes were sampled. For these lakes, density of stage V copepodids was used in data analysis because we were able to infer the species of diaptomid from the body length of the copepodids.

Mean density of adult diaptomids at each sampling occasion was calculated from the replicate vertical tows. Analysis was conducted using the maximum of the mean densities of each species in each year. If a lake was sampled over several years, maximum annual densities were averaged over all sample years.

Lake depth was determined with a handheld sonar gun (Manta FR-100) along transects parallel and perpendicular to the long axis of a lake. Surface area was estimated by digitizing 7.5 min USGS topographic maps. Elevation was determined from these maps.

Each time a lake was sampled, water temperature (°C) was measured at 1 m intervals from the lake surface to 30 m over the deepest point in the lake using an Omega 871A digital thermo-couple. Since very few of the lakes in this study were thermally-stratified, water temperature at a depth of 1 m was used in data analysis. Water samples were taken from a depth of 1 m over the deepest point in a lake using a 1.5 l Van Dorn bottle. Frozen filtered and unfiltered water samples were transported to the Cooperative Chemistry Analytical Laboratory at Oregon State University for chemical analysis (total Kjeldahl nitrogen, total phosphorus, ortho phosphorus-P, nitrate-N, ammonia-N, pH, alkalinity, conductivity). For each lake, values for each of these variables were averaged first for each year and then over all years that the lake was sampled.

To assess fish effects on large copepods, lakes were divided into three categories: lakes with high trout densities, lakes with low trout densities, and lakes with no trout. Lakes were placed in these categories based on mark-recapture estimates and information from stocking records. Absence of fish from lakes was verified by gill-netting, angling, and snorkeling.

In nine lakes, fish density was estimated by mark-recapture (Liss et al. 1995; Gresswell et al. 1997). Fish were captured by angling with lures and artificial flies, fm-clipped, and released. Injured fish (e.g., bleeding from gills or tongue) were not marked. Post-release mortality was probably low because only one of 24 marked fish died after being held overnight in enclosures. Recapture with variable mesh gill nets usually began the day following completion of marking. Abundance was estimated by a single-census Petersen estimator and 95% confidence limits were calculated (Ricker 1975). Both angling and gill-netting tended to under-sample small fish. Consequently, abundance estimates pertain to fish 177 mm total length. Trout this size are about 2-3 years old (Liss et al. 1995).

Some NOCA lakes in which trout do not reproduce are periodically stocked with fly at low densities. Lack of reproduction was determined from NOCA records (Jarvis 1987) and verified by field observations including lack of suitable spawning habitat, failure to observe fry or smaller fish, and little variation in age and size structure of captured fish. In NOCA, the average density of trout fry stocked in higher-elevation lakes (= 1100 m) from 1976 to 1993 was 179 fish ha^-1 (range 59.8-375 fish ha^-1; N=37) and the average interval between stocking was >5 years (unpublished records, North Cascades National Park Service Complex, 2105 Highway 20, Sedro Wooley, WA, 98284).

Lower elevation lakes (<1100 m) in NOCA are more accessible to anglers and tend to be more heavily stocked than higher elevation lakes. Fry have been stocked at high densities in two small, low elevation (1031 m) lakes, Upper Panther Lake (surface area = 0.1 ha, mean stocking density = 833 fry/ha, range = 500-1000 fly/ha, mean stocking interval = 3.5 years, N =3) and Lower Panther Lake (surface area = 0.2 ha, mean stocking density = 933 fry/ha, range = 750-1500 fly/ha, mean stocking interval = 3.7 years, N = 4). Upper Panther Lake is located a few meters away from Lower Panther Lake. In late summer 1990 fish were removed from both lakes by intensive gill-netting and angling. Trout density in these lakes was low at the time of fish removal. The lakes are small and clear and absence of trout following removal procedures was verified by observation from shore and snorkeling. In fall 1990, Lower Panther Lake was stocked with fly (O. clarki) at a relatively high density (750 fly/ha) whereas Upper Panther Lake was not stocked with fish. Zooplankton and abiotic factors were sampled in Upper and Lower Panther Lakes prior to fish removal in 1990 and in each subsequent summer from 1991 through 1993. In 1992, trout densities in Lower Panther were estimated by mark-recapture.

Larval salamanders (Ambystoma macrodactylum and A. gracile) were present in many fishless lakes (Liss et al. 1995; Tyler et al. 1998). Salamander predation can influence the structure of zooplankton communities (Dodson 1970, 1974; Sprules 1972; Giguere 1979; Morn 1987). Larval salamander densities were assessed by snorkeling the perimeter of lakes and carefully searching through bottom materials (Liss et al. 1995; Tyler et al. 1998).

Data Analysis

Correlations between densities of diaptomid species and abiotic factors were investigated by determining Pearson Correlation Coefficients and testing the coefficients for significance ( = 0.05). To assess the influence of abiotic factors on presence-absence of large and small diaptomid species, lakes were grouped into three categories: lakes with only large diaptomids, lakes with only small diaptomids, and lakes with both large and small diaptomids (Table 4.1). To determine if levels of abiotic factors differed among the three categories, pairwise comparisons of each abiotic factor between categories were conducted using the Kruskall-Wallis test. A Bonferroni adjustment (Miller 1981) was used to maintain the Type I error rate at 0.05 for the three pair-wise comparisons of each abiotic factor. Each comparison was judged to be significantly different if P 0.017 (0.05 ÷ 3). We also compared levels of abiotic factors between lakes with the small diaptomid (a group composed of lakes with only the small diaptomid and lakes with both large and small diaptomids) and lakes where the small diaptomid was absent (lakes with only large diaptomids) using the Kruskall-Wallis test ( = 0.05).

To determine the relative effects of abiotic factors and large diaptomids on small diaptomid density we fitted a general linear regression model with the logarithm of small copepod density as the dependent variable and selected abiotic factors, large diaptomid density, and lake category (i.e., lakes with only large diaptomids, lakes with only small diaptomids, and lakes with both large and small diaptomids) as independent variables. We compared the slopes of the relationships between the dependent and independent variables for each lake category.

To determine if large copepod density differed statistically between the three trout density categories (i.e., high trout density, low trout density, no fish), pairwise comparisons of large copepod densities between the categories were conducted using the Kruskall-Wallis test. Again, to maintain the Type I error rate at 0.05 we used a Bonferroni adjustment and individual comparisons were judged to be significantly different if P 0.017.

Interannual trends in the densities of large copepods in both Upper and Lower Panther Lakes were assessed using regression analysis. To determine if large diaptomid density differed between Upper Panther Lake and Lower Panther Lake, we compared diaptomid densities between the two lakes using the Kruskall-Wallis test ( = 0.05).