Crustacean zooplankton are important components of the pelagic community in high mountain lakes. The distribution and abundance of crustacean zooplankton in high elevation lakes is influenced by abiotic factors including temperature (Anderson, 1971; Moore, 1978), lake morphometry (Stoddard, 1987), and lake chemistry (Anderson, 1974; Stoddard, 1987) and biotic factors particularly invertebrate (Sprules, 1972; Dodson, 1974; Giguere, 1979; Neill, 1989) and vertebrate (Sprules, 1972, Dodson, 1974: Giguere, 1979, Anderson, 1980; Stoddard, 1987) predation

We hypothesized that effects of vertebrate predation on crustacean zooplankton would vary along a gradient of increasing predator density with the greatest effects occurring at high vertebrate predator densities. Vertebrate predators may prey selectively on larger species causing the zooplankton to be dominated by smaller-bodied froms (Brooks and Dodson, 1965; Zaret, 1980). The purpose of this section is to examine the relative importance of abiotic conditions and biotic factors including predation by invertebrates, salamanders, and fish in determining the distribution and abundance of crustacean zooplankton in NOCA lakes.

Crustacean zooplankton were sampled in 53 lakes from 1989-1993 (Table 1). In each lake, three replicate vertical tows and a horizontal tow were taken with a 20 cm diameter number 25 (64 µm mesh) zooplankton net, except in 1989 when only one vertical tow was taken. The vertical tows were taken near the deepest part of the lake. The net was lowered to within one meter of the lake bottom and towed upward at a rate of about 0.5 m/sec. The horizontal tow was taken by towing the net nearshore with an inflatable boat. 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. The split portion was poured into a settling chamber and left to settle for 24 hours. Organisms were counted using an inverted microscope at 100X magnification. Average body length of adult female zooplantkers of each taxon was measured.

Mean density of adult crustacean zooplankton at each sampling occasion was calculated from the replicate vertical tows. If a lake was sampled more than once per year, mean density for each taxon for each year was calculated. If a lake was sampled over several years, densities were averaged over all sample years. Thus, in each lake, each taxon was represented by a single mean density. Data from the horizontal tows are difficult to accurately quantify and were used only to assess presence of species in a lake.

Methods for sampling abiotic variables [elevation (elev), maximum depth, surface area, water temperature at a depth of one meter, alkalinity, conductivity, pH, total Kjeldahl nitrogen (TKN), nitrate (NO₃), ammonia (NH₃), total phosphorus (TP), and ortho-phosphorus (OP)] were described in previous sections of this report (see Lake Classification and Rotifer section). 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. Sampling procedures for salamanders were described in the Amphibian section of this report.

Lakes with fish were grouped into two categories: lakes with non-reproducing fish and lakes with reproducing fish. In general, naturally reproducing fish populations have a more complex size and age structure and higher densities than do non-reproducing fish (see Fish section). Densities of fish > 170 mm length in eight lakes with reproducing fish were estimated by mark-recapture and ranged from 450-600 fish/ha with the exception of MR16 (92 fish/ha). In lakes without fish reproduction, average density of stocked trout fry from 1976 to 1993 was 179 fish/ha (range 59.8-375 fish/ha; N=37), except for Thunder, Lower Panther, and Upper Panther lakes which are relatively heavily stocked (data provided by Reed Glesne, North Cascades National Park Service Complex, 2105 Highway 20, Sedro Woolley, WA, 98284).

Data Analysis

The relative occurrence of crustacean zooplankton taxa among lakes in NOCA was assessed by determining the proportion of lakes in which each taxon occurred and by calculating niche breadth, B,:

where P_ijis the proportion of individuals of the i-th species found in the sample from the j-th lake, and R_ijis the summation of the proportions of the i-th species over all lakes. The variable B can range from 1, when a species is present in only one lake, to 53 when a species is equally common in all lakes.

Major kinds of crustacean zooplankton communities were identified with hierarchical agglomerative cluster analysis (Ludwig and Reynolds, 1988) based upon the average proportion of each taxon in each lake. The fusion strategy was the unweighted pair-grouping method (group average) and the resemblance measure was the Euclidean distance. Species diversity for each lake cluster was expressed by Shannon's information measure (Shannon and Weaver, 1949):

where n_i is the number of organisms belonging to the i-th taxon in a sample of N individuals. Similarity in taxonomic composition among clusters was expressed as Percentage Similarity (Ludwig and Reynolds, 1988):

where p_ij and p_ik are the mean proportions of species i in samples belonging to clusters j and k, respectively.

An ordination of the proportional species abundance data was performed using detrended correspondence analysis (DCA, Hill and Gauch, 1980). Principal Components Analysis (Ludwig and Reynolds, 1988) was performed on abiotic variables. To relate variation between lakes along the DCA axes to abiotic variables, DCA axis scores were regressed against the first four principal components (only the first four components has eigenvalues >1.0). A t-test was used to determine if each of the standardized partial regression coefficients was significantly different from zero ( = 0.05).

The Wilcoxon rank sum test (Snedecor and Cochran, 1989) was used to determine if lakes with reproducing fish differed statistically from lakes with non-reproducing fish in average densities (no/l) of the copepods Diaptomus kenai and Diaptomus tyrrelli and in average values of each abiotic variable. Relationships among D. kenai and D. tyrrelli and abiotic variables were investigated by determining Pearson Correlation Coefficients and testing the coefficients for significance using a t-test. To investigate patterns of co-occurrence of D. kenai and D. tyrrelli we compared the proportion of lakes in which the two species co-occurred with the proportion in which they did not co-occur (t-test, Steel and Torrie, 1980, pg. 482).

Crustacean Zooplankton Communities

Crustacean zooplankton communities were aggregated into eight clusters (Figure 1, Table 2). A single species was dominant (>50% proportional abundance) in each cluster. In some clusters single-species dominance was high (>80% proportional abundance, clusters I and V). Several subdominant species (>5% proportional abundance) were usually present in each cluster. Percentage similarity of clusters was low (Table 3), indicating a high degree of uniqueness in taxa composition and proportional abundances among clusters.

Figure 1. Cluster analysis of 53 lakes based on proportional abundance of crustacean zooplankton species, North Cascades National Park Service Complex.

Table 2. Lake clusters based on proportional abundance of crustacean zooplankton species in North Cascades National Park Service Complex. Dominants are species with 50% proportional abundance. Subdominants are species with 5% proportional abundances. Average number of species in each cluster is in parenthesis.

Cluster Type	Number of Lakes	Dominant Species	Average Proportion of Dominant	Subdominant Species	Average Proportion of Subdominants	Species Diversity (H)	Lakes
I	20	Diaptomus kenai	0.89	Holopedium gibberum Daphnia	0.05 0.05	0.52 (2.3)	BEAR, MR15-1, MM11, MR16, REVL, MA2, TAP1, VULC, WILD, MR13-2, TAP4, TAP2, EGG, MR13-1, WADD, MR11, MONO, PM5-3, SKYM, COPP
II	10	Daphnia rosea	0.60	Diaptomus kenai Bosmina longirostris Diaphanosoma brachyurum	0.12 0.06 0.05	1.49 (6.7)	BATT, HOZO, THUN, DAGG, LS2, NERT, MR12, LS1,LS3, PANL
III	7	Holopedium gibberum	0.75	Diaptomus kenai	0.19	0.78 (4.1)	BOUC, RAIN, DOUB, THRM, EP6, THRL, MCAL
IV	7	Diaptomus tyrrelli	0.68	Daphnia rosea Cyclopoid copepods Ceriodaphnia quadrangula	0.08 0.06 0.06	1.33 (5.1)	JUAN, MR2, KETT, MR9, TRIU, MR3, TRIL
V	4	Diaptomus arcticus	0.80	Chydorus sphaericus Diaptomus kenai	0.11 0.09	0.66 (2.0)	GNVW, TRAP, MA3, WADDM
VI	3	Diaphanosoma brachyurum	0.62	Diaptomus kenai Bosmina longirostris Daphnia rosea	0.13 0.11 0.10	1.25 (10.7)	PANU, RIDL, PYRA
VII	1	Diaptomus lintoni	0.56	Ceriodaphnia reticulata Daphnia rosea	0.24 0.20	1.03 (4.0)	COON
VIII	1	Daphnia middendorffiana	0.52	Diaptomus tyrrelli Diaptomus leptopus	0.35 0.13	0.35 (4.0)	WILL

Clusters are identified on the ordination in Figure 2. The first four PCA components accounted for 77.5% of the variation among abiotic factors. The first component was an expression of physical variables related to climate and certain chemical variables (Table 4). Temperature, pH, alkalinity, conductivity, and TKN had high positive loadings and elevation had high negative loading with the first component. For the regression of DCA axis 1 scores against PCA components, the standardized partial regression coefficient (b) of the first component was highly significant (b=0.615, t-test, P<0.0001) suggesting that abiotic factors that loaded heavily in this component could explain some of the variation among lake clusters along DCA axis 1 (Figure 2). The second PCA component was an expression of lake size (area, depth) and NO₃ and had a weak but significant negative relationship with DCA axis 1 (b=-0.219, t-test, P=0.038). The third component had high positive loadings of NH₃ and OP while the fourth component also heavily loaded OP. Neither the third nor the fourth component had a statistically significant relationship with DCA axis 1 (b=-0.142, P=0.174 and b=-0.043, P=0.678, respectively).

Figure 2. Ordination of lakes based on proportional abundance of crustacean zooplankton species in each lake. Lake clusters (Table 2) are shown. Arrows indicate direction of increase of abiotic variables along axis one.

None of the first four PCA components was significantly correlated with DCA axis 2. This means that variation along axis 2 cannot be related to the measured abiotic variables.

Cluster I

Cluster I (Table 2) represents the most common type of zooplankton community in the park. Thirty-eight percent of the surveyed lakes occurred in this cluster. This type of community is dominated by Diaptomus kenai, one of the largest diaptomid copepods (Table 5) and the most ubiquitous zooplankter in NOCA (Table 6). Proportional abundance of D. kenai is high (0.89) in these lakes and species diversity is low (H=0.52). Forty-five percent of the lakes in this cluster contained only D. kenai. The next most common species was Holopedium gibberum, followed by two species of Daphnia, D. pulex and D. middendorffiana. These species are the largest daphnids in the park but they are relatively uncommon (Table 6).

Most lakes in this cluster are relatively high elevation lakes. Ninety percent of the lakes occur in the alpine or subalpine vegetation zone on both the east and westslope of the Cascade range. Communities in these lakes are most similar to those in cluster III (Table 3), which is dominated by H. gibberum with D. kenai as the major subdominant species.

Lakes in this cluster are relatively deep and are characterized by low temperature, alkalinity, conductivity, TKN, and TP (Table 7). They appear to be relatively low productivity lakes.

Cluster II

Cluster II represents another common type of community (Table 2). Lakes in this cluster are dominated by Daphnia rosea, the most common daphnid and second most-common zooplankter in NOCA (Table 6). Diaptomus kenai is the major subdominant, but two species of cladocerans are also relatively common subdominants. Species diversity is the highest of all clusters (H=1.49).

Lakes in this cluster are mid-elevation lakes, although the lowest elevation lake in the park, Thunder Lake, occurs in the cluster. Ninety percent of the lakes occur in the forest zone. Temperature, alkalinity, conductivity, TKN, and TP are moderate relative to other clusters and nitrate concentration is low (Table 7). Lakes in cluster II are most similar to those in cluster VI. Both clusters have relatively high proportions of D. kenai and D. rosea.

Cluster Ill

This cluster is composed of large, deep, relatively high elevation lakes dominated by H. gibberum (Table 2). Holopedium gibberum is the third most-common crustacean zooplankton species among the surveyed lakes (Table 6). The subdominant species, D. kenai, maintains high relative abundance in these lakes. Lakes in this cluster are subalpine or high elevation forest lakes. Like cluster I, species diversity is low (H=0.78).

Lakes with H. gibberum and D. kenai as major co-occuring species are somewhat more common than indicated by cluster analysis. Cluster III is most similar to cluster I (Table 3), where D. kenai is the dominant species. Lakes with both species were placed in cluster I or Ill depending on the proportion of each of the two species. For nine of the 53 lakes (17%), the sum of the proportions of the two species was 0.90. Two of these lakes (PM 5-3 and M. Skymo) were placed in cluster I rather than Ill because the relative abundance of D. kenai was greater than H. gibberum.

The lakes in this cluster are characterized by high nitrate, but low temperature, alkalinity, conductivity, TKN, and TP and, like lakes in cluster I, they may be relatively less productive than lakes in other clusters.

Cluster IV

Lakes in this cluster (Table 2) are small, shallow, high elevation eastslope lakes (Table 7) dominated by the small-bodied diaptomid copepod Diaptomus tyrrelli (Table 5). Diaptomus tyrrelli maintains high densities in many of these lakes. Lakes in this cluster are either subalpine (6 lakes) or high elevation forest lakes. This is the only community type where cyclopoid copepods are relatively common. Chydorus sphaericus and the cladoceran predator Polyphemus pediculus are present in low abundance in a relatively large number of lakes. Dagger Lake maintains a high density of D. tyrrelli but was placed in cluster II rather than IV because the relative proportion of D. rosea (0.71) is greater than the proportion of D. tyrrelli. Species diversity in cluster IV is relatively high (H=1.33). Lakes in this cluster are characterized by moderate temperature, alkalinity, conductivity, TKN and high TP relative to other clusters (Table 7).

Cluster V

This cluster (Table 2), composed of only four lakes, is dominated by the copepod Diaptomus arcticus, the largest zooplankter in the park (Table 5). Proportional abundance of D. arcticus in these lakes is high (0.80). All lakes in this cluster are eastslope subalpine lakes. In NOCA, D. arcticus is relatively uncommon (Table 6) and was found in only 11% of surveyed lakes. Chydorus sphaericus and D. kenai are subdominant species. Like lakes in clusters I and Ill, species diversity is low (H=0.66).

Lakes in cluster V are large, deep (except M. Waddell) high elevation lakes (Table 7). The lakes are characterized by moderate temperature, alkalinity, conductivity, and TP, low TKN, and high concentrations of NO₃.

Cluster VI

This cluster is composed of three low elevation westslope forest lakes (Table 2 and 7). The lakes are dominated by the cladoceran Diaphanosoma brachyurum. Subdominant species, D. kenai, D. rosea, and Bosmina longirostris, are present in relatively high proportions. Cluster VI is most similar to cluster II (Figure 2, Table 3). The two clusters have the same species in common but the species differ in relative proportions between the clusters (Table 2). Cluster VI is characterized by relatively high temperature, conductivity, alkalinity, TKN, and TP (Table 7).

Clusters VII and VIII

These two clusters are each composed of a single low elevation forest lake (Tables 2 and 7). Cluster VII is an eastslope lake dominated by the large copepod Diaptomus lintoni. This copepod is extremely rare in NOCA high lakes (Table 2). It was found only in Coon Lake. Cluster VIII is a westslope lake dominated by the large daphnid, Daphnia middendorffianna (Table 5). Lakes in these two clusters may be some of the most productive lakes in NOCA. The two clusters have the highest average water temperature and TP of all lake clusters in NOCA (Table 7). TKN is also relatively high in both clusters. Cluster VIII has the highest alkalinity and conductivity of all clusters.

Patterns of Co-Occurrence of Vertebrate Predators and Crustacean Zooplankton

Distribution of Diaptomid Copepods

Five species of diaptomid copepods were found in NOCA. Three of these species, D. kenai, D. arcticus, and D. tyrrelli. are relatively common (Table 6). The other species, D. lintoni and D. leptopus, were found only in low elevation lakes. Diaptomus kenai and D. arcticus are the largest zooplankters in NOCA (Table 5) and are likely to be predacious (Zaret, 1980; Anderson, 1967, 1970, 1972). Diaptomus tyrrelli is a relatively small herbivorous copepod.

Diaptomus kenai is found in lakes at all elevations on both the east and westslope of NOCA (Table 8). In high elevation eastslope lakes, (>1600 m), D. kenai co-occurred with D. arcticus in four of the five lakes in which D. arcticus was found (Table 8). However, in eastslope lakes, neither large copepod commonly co-occurred with the small diaptomid, D. tyrrelli. In contrast, in low elevation westslope lakes, D. kenai and D. tyrrelli commonly co-occurred (Table 8).

Abiotic Factors and Invertebrate Predation

Abiotic factors may affect distribution and co-occurrence of large and small diaptomids. The density (no/I) of D. kenai was not significantly correlated with any of the measured abiotic variables (Table 9). The lack of correlation of D. kenai density with abiotic factors reflects the ability of this species to tolerate a relatively wide range of abiotic conditions in NOCA, which in part accounts for its status as the most ubiquitous crustacean zooplankter in high mountain lakes in the Complex (Table 6).

While density of D. kenai was not significantly correlated with any abiotic variable, density of D. tyrrelli had significant positive correlations with both TKN and TP (Table 9). Furthermore, many of the other abiotic variables were significantly correlated with TKN and TP (Table 10). In particular, TKN and TP were negatively related to lake depth.

*Statistically significant correlations.

Patterns of co-occurrence of D. kenai and D. tyrrelli are related to TKN, TP, and lake depth. Lakes where D. tyrrelli was allopatric or sympatric with D. kenai were relatively shallow lakes with moderate to high TKN ( 0.03 mg/I) and TP ( 0.004 mg/I) (Figures 3 and 4). Diaptomus kenai occurred, either allopatrically or sympatrically with D. tyrrelli, over the entire range of lake depths and concentrations of TKN (Figure 3), but was absent from lakes with very high TP (TP0.1 1, Figure 4).

Figure 3. Relationships between the concentration of Kjeldahl nitrogen and lake depth relative to occurrence of D. keni, D. tyrrelli, and lakes inhabited by both species.

Figure 5. Relationships between the concentration of total phosphorus and lake depth relative to occurrence of D. keni, D. tyrrelli, and lakes inhabited by both species.

Although the densities of D. kenai and D. tyrrelli were not significantly correlated with water temperature (possibly due to reduced densities of both species at the extremes of their temperature ranges), the two species may differ in temperature tolerance. Diaptomus kenai rarely occurred in lakes with an average water temperature over 15-16°C (Table 11). It occurred at very low densities in Thunder Lake (21°C) in late spring before the water warmed. It was not found in this lake over the summer. Diaptomus kenai density was low in Hozomeen Lake (0.03/l, 17.5°C) however, this lake also supports reproducing trout. Diaptomus kenai was found only in horizontal tows in Ridley Lake (17.9°C), indicating that its density in this lake is very low (<0.001/I). In Figure 4, the two lakes with the highest TP concentrations are both very warm (Juanita, 19.7deg.C; Willow, 19.5°C) Perhaps D. kenai does not occur in these lakes because of their high temperature.

*Ridley Lakes = 17.9°C

In contrast to D. kenai, Diaptomus tyrrelli is able to maintain high densities in lakes with relatively high water temperatures (e.g., Juanita, Lower Kettling, MR2, MR3) (Table 11). Thus, D. tyrrelli may be capable of tolerating somewhat warmer water than D. kenai, while D. kenai (and D. arcticus) may be able to withstand colder water than D. tyrrelli. However, there is considerable overlap in the temperature ranges of all these species (Table 11).

Abiotic factors alone are insufficient to completely explain the distribution and abundance of D. tyrrelli. Of the 27 lakes with abiotic conditions within the range that D. tyrrelli occurs, the small copepod is absent from 11 lakes (40%; absent from 38% if the lower limit of TP=0.007; Table 12). Seven of the 11 lakes are high elevation eastslope lakes.

Perhaps predation on the small copepod by large copepods, particularly the ubiquitous D. kenai, may be responsible, in part, for absence of D. tyrrelli from some of the lakes with abiotic conditions apparently favorable for this species. Diaptomus tyrrelli maintains higher densities when it is allopatric than when it co-occurs with large copepods (Table 12). Most of the lakes where the small copepod is allopatric are high elevation eastslope lakes. In these lakes, the density of large copepods was lower in lakes where they co-occurred with D. tyrrelli than in lakes where they were allopatric (Table 12), possibly suggesting that the smaller copepod is able to colonize and increase in density when the density of larger copepods is reduced.

Table 12. Density of D. tyrrelli and large copepods in lakes where TKN 0.04, TP 0.004, and temp 11.4°C.

	D. tyrrelli Absent	D. tyrrelli Allopatric	D. tyrrelli Sympatric
Large copepod density
Eastslope Westslope All lakes	1.15 (0.00-2.42) 0.36 (0.09-0.81) 0.88 (0.00-2.42)	0.00 0.00 0.00	0.18 (0.01-0.32) 1.06 (0.03-2.11) 0.75 (0.01-2.11)
D. tyrrelli density
Eastslope Westslope All lakes	0.00 0.00 0.00	7.95 (0.13-20.14)   2.042.64 (0.13-20.14)	0.73 (0.41-0.95) 0.17 (0.11-0.36) 0.27 (0.001-0.95)
N
Eastslope Westslope All lakes	7 4 11	6 1 7	3 6 9

Although co-occurrence of the two species was common in low elevation westslope lakes, in lakes where the two species were sympatric, D. kenai maintained higher densities than did D. tyrrelli (Table 12). In the only low elevation lake where it was allopatric (Willow Lake), D. tyrrelli density was considerably higher than in low elevation lakes where it was sympatric with D. kenai. Thus, although the two species are capable of co-occurring in low elevation westslope lakes, D. kenai may still depress the density of D. tyrrelli density.

Vertebrate Predators

At present, analysis of the potential effects of vertebrate predators has been conducted only for eastslope lakes. Vertebrate predators are common in eastslope lakes. Of 26 lakes surveyed for vertebrate predators (see Fish and Amphibian sections), 21 supported either fish or salamander (A. macrodactylum) larvae (Table 13). In lakes with fish, larval salamander densities were very low. In some fishless lakes, salamander larvae were very abundant.

* CT=cutthroat trout, RB=rainbow trout
** number of larvae/100 m shoreline determined by snorkel survey (see Amphibian section for sampling techinque).

A few lakes had no vertebrate predators. The reason salamanders are absent in these lakes is not yet clear. Most of the lakes lacking salamanders are relatively cold. Two lakes, Upper and Lower Silent, are the highest elevation lakes on the eastslope in NOCA, occurring at over 2000 m. Their watersheds, and that of Greenview, the coldest lake without vertebrate predators (9.1°C) are almost entirely rock and snow and may be unfavorable habitat for metamorphosed A. macrodactylum.

Both D. kenai, a large copepod, and D. middendorffianna, the largest daphnid in NOCA occur in lakes with low densities of salamander larvae (Table 14). In contrast the small copepod, D. tyrrelli, and small cladocerans occur in lakes where larval salamander density is high. However, there are important exceptions to the putative association of small-bodied crustaceans with high densities of larval salamanders. D. tyrrelli occurs in Juanita Lake (Table 8) but larval salamander density in this lake in 1993 was relatively low (Table 13). Although Juanita occurs at high elevation (2033 m), it is a shallow lake perched on a ridgeline and water temperature can be rather high (19.7°C) perhaps excluding D. kenai and allowing D. tyrrelli to successfully colonize. In MR12, salamander density is very high (Table 12), but D. kenai is present, although at relatively low density (0.27/I). Further analysis of the relationship of crustacean zooplankton and salamander larvae is necessary.

* 1990, prior to stocking fish

Examination of the effects of fish predation on large copepods in eastslope lakes is hindered by the scarcity of lakes with no vertebrate predators (Table 13). Adult copepods were found in three of the five eastslope lakes without vertebrate predators.

The copepods in these lake were all large-bodied species but their densities were low. At this point in our analysis it is not clear why densities of large copepods are low in these lakes, however, low water temperature may be a factor.

The density of large copepods in lakes with reproducing fish was significantly lower than in lakes with non-reproducing fish (Table 15, Wilcoxon test, P=0.01). When MR16, a lake with a low density of reproducing fish, was omitted from the reproducing trout category and added to the non-reproducing trout group (essentially resulting in comparison of copepod densities between lakes with high and low fish densities), the difference in large copepod densities between lakes with reproducing fish (excluding MR16) and the group of lakes including non-reproducing fish and MR16 was highly significant (P<0.001).

^aLow fish density

There was no significant difference in the density of D. tyrrelli between lakes with reproducing and non-reproducing fish (Wilcoxon test, P>0.2). However, D. tyrrelli was present in five of the ten lakes with reproducing fish. In 2 of the 5 lakes (Doubtful, Trapper) with reproducing fish where the small copepod is absent, abiotic conditions were not within the range of conditions that D. tyrrelli occurs. Diaptomus kenai density is high in MR 16, perhaps preventing D. tyrrelli from colonizing. It is unclear why the small copepod is absent from Battalion and Rainbow Lakes. Rainbow is relatively large (6.3 ha and deep (10.4 m). All crustacean zooplankton are scarce in Battalion Lake. In four of the five lakes with non-reproducing fish, abiotic conditions were not within the range of conditions that the small copepod occurs.

The elevation of lakes with non-reproducing trout was significantly higher than the elevation of lakes with reproducing trout (Table 16, Wilcoxon test, P=0.05). There were no statistically significant differences in other abiotic factors between lakes with reproducing and non-reproducing fish (Table 14). It is unlikely that the difference in elevation between lakes with reproducing and non-reproducing fish can account for the difference in densities of large copepods. Diaptomus kenai density was not significantly correlated with elevation (Table 9). Furthermore the elevation ranges of both lakes with reproducing fish and lakes with non-reproducing fish was well within the elevation zone where both species of large copepods occur. Average water temperature was not significantly different between lakes with reproducing and non-reproducing fish. Thus, relative to lakes with low densities of fish, high densities of reproducing trout can reduce or eliminate large copepods from lakes. However, we cannot conclude that non- reproducing fish have no effect on large copepods in eastslope lakes since we do not know at what densities large copepods would occur if lakes with non-reproducing fish were fishless.

*Statistically significant difference between lakes with reproducing and non-reproducing fish.

Several different types of crustacean zooplankton communities were identified in NOCA. Variation in physical and chemical conditions in lakes due principally to variation in elevation, aspect, extent of development of soils and vegetation, lake morphometry, and, to a some extent, lithology create different environments suitable for different suites of species. High elevation westslope lakes belonged predominantly to clusters I and Ill (Figure 5). The dominant species in these clusters were D. kenai and H. gibberum. These lakes were characterized by low average water temperature and low alkalinity and nutrients (TKN, TP). The lakes may be poorly productive relative to other NOCA lakes. D. kenai and H. gibberum are apparently well-suited to these conditions.

High elevation zooplankton communities are more diverse on the eastslope than on the westslope (Figure 5). Communities belonging to clusters I and Ill also were present on the eastslope. Cluster V, dominated by D. arcticus, was found only on the eastslope. Lakes in cluster V are deep lakes and, like clusters I and Ill, they may be lower productivity lakes characterized by low water temperature, pH, alkalinity, and nutrients. Cluster IV also occurred only on the eastslope. Lakes in this cluster, dominated by the small copepod D. tyrrelli, were smaller, shallower lakes

Low elevation lakes tend to have higher water temperatures, alkalinity, conductivity, and concentration of nutrients than high elevation lakes, with the exception of lakes in cluster IV. Zooplankton communities belonging to clusters II and VI were the major kinds of communities in low elevation westslope lakes (Figure 5). These clusters share similar species and are characterized by similar abiotic conditions.

Figure 5. Crustacean zooplankton communities at high and low elevation on the east and westslope of the crest of the Cascades, North Cascades National Park Service Complex. Dominant species indicated by an *.

Distribution and patterns of co-occurrence of large and small copepods in NOCA is related to abiotic conditions including lake morphometry, water chemistry, and possibly water temperature, and to biotic factors especially vertebrate and invertebrate predation (Figure 6). The large copepod Diaptomus kenai is widespread in NOCA and is apparently able to tolerate a wide range of abiotic conditions. However, its abundance and distribution may be limited by water temperature (range of occurrence approximately 9-10°C to 16-17°C) In eastslope lakes the large copepod D. arcticus occurs over a more limited temperature range(9.1-13.8°C) than D. kenai, suggesting that it may be less capable than D. kenai of tolerating relatively high temperatures.

Figure 6. Conceptual model of the possible interactions of lake morphometry, water, temperatue, water chemistry, and invertebrate and vertebrate predation in determining distribution and abundance of diaptomid copepods in NOCA.

In NOCA D. tyrrelli is apparently not able to tolerate as wide a range of abiotic conditions as D. kenai. Diaptomus tyrrelli appears to prefer lakes with higher water TKN (0.03 mg/I) and TP (0.008 mg/I). In NOCA, lakes with these characteristics are usually relatively shallow ( 5-6 m maximum depth). Undoubtedly D. tyrrelli is absent from many lakes due to unfavorable abiotic conditions (low temperature and concentrations of TKN and TP) and high water temperature may limit the distribution of large copepods. However, the abiotic factors we measured alone are not sufficient to completely explain patterns of distribution and abundance of D. tyrrelli or large copepods. Diaptomus kenai and D. tyrrelli are absent from many lakes with abiotic conditions that are apparently suitable for these species. Diaptomus kenai distribution and abundance appears to be influenced by the level of vertebrate predation, particularly fish predation (Figure 6). Reproducing populations of trout that reach high densities can reduce D. kenai density and apparently eliminate the species from some eastslope lakes. Anderson (1967) noted the disappearance of large crustacean zooplankton, including D. arcticus and Daphnia middendorffiana, from lakes in the Canadian Rockies following introduction of trout. Stoddard (1987) documented the absence of large-bodied crustacean zooplankton including large diaptomids, from mountain lakes in the Sierra Nevada.

In eastslope Lakes in NOCA, D. tyrrelli rarely co-occurs with large copepods and when large and small diaptomids do co-occur, the density of large copepods is usually low. Although co-occurrence of large and small diaptomids in low elevation westslope lakes is common, the density of D. tyrrelli in lake with D. kenai is depressed relative to its density in lakes without the large copepod. Possibly, predation by large copepods can reduce D. tyrrelli density or eliminate the small copepod from some lakes. Anderson (1967) documented the appearance and subsequent increase in abundance of D. tyrrelli in lakes in the Canadian Rockies after large diaptomids had been eliminated by fish introductions. Interestingly, he also noted an increase in densities of cyclopoid copepods, presumably from feeding on abundant juvenile stages of D. tyrrelli. Cyclopoids were most abundant in lakes with high D. tyrrelli densities in NOCA.

In NOCA, if large copepod density is reduced or large copepods are eliminated by either high temperature or high fish (and perhaps larval salamander) density, D. tyrrelli may colonize and become abundant. Whether the small copepod successfully colonizes may depend on water temperature and nutrient conditions (Figure 6).

Cluster analysis showed that lakes with D. tyrrelli as the dominant species (cluster IV) composed an important type of crustacean zooplankton community in NOCA. Moreover, in cluster VIII (Willow Lake) D. tyrrelli was an important subdominant species. The second axis of the ordination of lakes separated clusters IV and VIII (Willow Lake) from the other cluster and was not significantly correlated with any of the measured abiotic variables. Variation among lakes along the second axis on the ordination may be related in part to high vertebrate predation and possibly high temperature (Willow Lake).