The taxonomic structure of phytoplankton assemblages in lakes located in mountainous regions and lakes at high latitudes have been shown to correspond to a variety of environmental, physical, and chemical variables such as climate, lake depth, lake area, water clarity, water quality, nutrients, and lake productivity (Pechlaner, 1971; Ilmavirta et al., 1984; Arvola, 1986; Earle et al., 1986; Earle et al., 1987; Knoechel and Cambell, 1988; Pinel-Alloul et al., 1990). Lakes in NOCA vary in water quality among the four main vegetation zones (alpine, subalpine, high forest and low forest). For example, alpine lakes have low water temperature and high concentrations of nitrate- nitrogen relative to low-forest lakes. Decreasing elevation and associated changes in vegetation zone tend to correspond with increases in water temperature, pH, alkalinity, conductivity, concentration of total nitrogen, and concentration of ammonia-nitrogen. Given this array of lake characteristics, a reasonable hypothesis was that characteristics of phytoplankton assemblages in NOCA lakes would correspond with the variation in physical and chemical conditions that exists among lakes in different vegetation zones. The objectives of this study were to:

(1) document the characteristics of phytoplankton assemblages of NOCA lakes;

(2) examine relationships between phytoplankton assemblages and the physical and chemical features of the lakes.

An additional objective was to assemble a phytoplankton data base to assess correspondence between the characteristics of phytoplankton and zooplankton communities. The third objective was addressed in the crustacean zooplankton section of this report.

From June 1988 through September 1993, 182 phytoplankton samples were collected from 64 lakes in NOCA (Table 1). Phytoplankton samples were collected throughout the park from lakes in vegetation zones ranging from low-elevation forest to alpine. Most of the samples were collected between July and September (Table 1).

All samples were collected using a Van-Dorn-style water sampler from a depth of 1 m near the deepest portion of each lake. Each 1-liter phytoplankton sample was preserved in Lugol's solution. The phytoplankton in each sample were enumerated using an inverted microscope (Mclntire et al., 1993). In nearly all cases, 500 cells were counted per sample. Counts of only 100 cells were made in five of the samples because of the presence of large amounts of debris from glacial out-wash. A multiplier was estimated for each taxon based on cell diameter and was used to transform count data into estimates of cell biovolume.

Analysis of the phytoplankton data was conducted in two parts. The first was an analysis at the division level, or in the case of Chrysophyta, the class level of classification using data collected from 1988 through 1993. Acronyms for the division and class levels of classification were as follows: Chlorophyta (CHL), Chrysophyta (CHR), Bacillariophyta (BAC), Cyanophyta (CYN), Pyrrhophyta (PYR), Cryptophyta (CRY), Euglenophyta (EUG), and unknown (UNK). The second was an analysis of the phytoplankton at the lowest possible taxonomic level (usually at the species level) and included only the data collected in 1989. Ordinations of the data at the division and species levels were performed using detrended correspondence analysis and the program DECORANA (Hill and Gauch, 1980). Proportional abundances and biovolumes were calculated using the program AIDN (Overtone et al., 1987). A cluster analysis (Mclntire, 1973) was used to group the samples in ordination space. The SYSTAT (Wilkinson, 1987) program, MGLH, was used for multiple regression tests. Correlation analyses were run on the SYSTAT program, CORR.

Community Characteristics: 1988 - 1993

The phytoplankton samples processed during the six-year study contained a total of 153 taxa, which included 52 chlorophytes (Chlorophyta), 43 chrysophytes (Chrysophyceae), 28 diatoms (Bacillariophytceae), 18 cyanobacteria (Cyanophyta), 4 dinoflagellates (Pyrrhophyta), 4 cryptomonads (Cryptophyta), 1 euglenoid (Euglenophyta), and 3 unknown taxa (Table 2). Collectively for all samples with 500 cell counts, Chlorella sp. and Aphanocapsa delicatissima had the highest average proportional abundances at 0.171 and 0.156, respectively (Table 3). Twenty other taxa had average proportional abundances of between 0.01 and 0.06. The remaining 131 taxa (data not shown) had average proportional abundances of less than 0.01. The average proportional abundances of the taxa increased when the averages were calculated using data only from the lakes in which each taxon was found (Table 3).

¹CHR = Chrysophyta
CHL = Chlorophyta
BAC = Bacillariophyta
CRY = Cryptophyta
CYN = Cyanophyta
EUG = Euglenophyta
PYR = Pyrrhophyta
UNK = Unknown

Table 3. Average proportional abundances ( 5%) for phytoplankton taxa in the 178 samples collected during the period from 1988 to 1993, number of samples in which each taxon was present, and average proportional abundance in all samples in which each taxon was present.

Taxa	Number of Samples Taxa Present	Average Proportional Abundance
		All Samples	Samples Present
Chlorella sp.	87	.171	.348
Aphanocapsa delicatissima	72	.156	.385
Chromulina parvula	62	.060	.170
Synechocystis sp.	65	.043	.116
Chromulina sp.	63	.041	.119
Ochromonas pinguis	42	.039	.163
Unidentified chrysophyte1	64	.033	.091
Aphanothece clathrata	19	.031	.289
Diogenes sp.	27	.031	.199
Ochromonas sp.	58	.031	.094
Chrysophyte statospore2	49	.023	.064
Unknown chrysophyte3	65	.021	.058
Chroococus minimu	17	.018	.189
Oocystis solitaria	42	.016	.068
Ochromonas ovalis	18	.016	.155
Unidentified4	23	.015	.114
Gleocystis planctonica	16	.014	.150
Rhodomonas sp.	51	.013	.040
Gymondinium-like	37	.013	.063
Unknown chlorophyte5	18	.011	.109
Chrysophyte statospore6	33	.010	.053
Pseudokephyrion sp.	36	.010	.051

¹Taxon I.D. No. 114
²Taxon I.D. No. 107
³Taxon I.D. No. 100
⁴Taxon I.D. No. 141
⁵Taxon I.D. No. 126
⁶Taxon I.D. No. 110

Collectively for all samples with 500 cell counts, cyanobacteria, chrysophytes, and chlorophytes were the dominant organisms (Table 4). Cyanophyta had the highest proportional cell densities, whereas the proportional cell biovolumes ere dominated by the Chrysophyta. Although low in number in the counts and proportional cell densities, dinoflagellates had the second highest proportional cell biovolume at 0.157.

The sample data were separated into alpine, subalpine, high-forest and low-forest vegetation zones based on the lake classification (Lomnicky et al., 1989; Liss et al., 1991). Subalpine and high-forest lakes had the largest number of taxa per sample, and alpine lakes had the smallest number (Table 5). However, taxa from each taxonomic division and class were found in each vegetation zone (Table 5). Alpine lakes contained a fairly even number of diatom, chlorophyte and chrysophyte taxa. Subalpine lakes and high-forest lakes were dominated by chrysophyte and chlorophyte taxa, where as chrysophyte taxa dominated low-forest lakes. The number of diatom taxa per sample was greatest in subalpine lakes, and the number of chlorophyte taxa was greatest in subalpine and high-forest lakes. The number of chrysophyte taxa per lake was highest in subalpine, high-forest and low-forest lakes. The numbers of Cryptophyta, Cyanophyta, Euglenophyta, and Pyrrhophyta taxa per lake were low, but evenly distributed among the four vegetation zones.

Representatives of the chlorophytes and chrysophytes were the dominant taxa in alpine lakes based on 500 cell counts (Table 6). Chlorophytes, chrysophytes, and cyanobacteria were the dominant taxa in subalpine lakes and high-forest lakes. Cyanobacteria were the dominant taxa in low-forest lakes. Similar results were found for proportional cell densities, except that cyanobacteria taxa dominated both high- forest and low-forest lakes (Table 6). Chrysophyte taxa dominated the proportional biovolumes of all lake types (Table 6). Dinoflagellate biovolumes were lowest in low-forest lakes, whereas diatom, chlorophyte and cryptophyte biovolumes were highest in alpine lakes. Taxa in the other divisions did not contribute substantially to the phytoplankton biovolumes of any lake type.

Vegetation zones were distributed in ordination space relative to the distribution of the organisms by taxonomic division or class to relate the scores of the first two axes to environmental and water quality variables (Fig. 1). The distribution of the phytoplankton divisions and classes corresponded to the distribution of the vegetation zones relative to the maximum proportional abundances of each division or class (Table 6). Although low in abundance, Cryptophyta (CRY) and Bacillariophyta (BAC) were at maximal proportional abundances in alpine lakes, whereas Cyanophyta (CYN) were abundant in low-forest lakes. The other divisions or classes (Euglenophyta and unknowns were deleted because they were rare) had maximal abundances in subalpine and alpine lakes. None of the divisions or classes had maximum proportional abundances in high-forest lakes. Increasing scores on axis one corresponded to increasing elevation and concentration of nitrate-nitrogen and decreasing water temperature, pH, alkalinity, conductivity, concentration of Kjeldahl-nitrogen, and concentration of ammonia-nitrogen (Figs. 2-4). Increasing scores on axis two corresponded to increasing concentrations of total phosphorus (Fig. 4).

Community Characteristics: 1989

Lake ordinations corresponding to species ordinations (data not shown) were used to relate the axis scores to environmental and water quality variables for samples collected in 1989 (Fig. 5). Increasing scores along axis one corresponded to increasing concentrations of Kjeldahl-nitrogen and a transition in vegetation zones from alpine to low-forest (multiple regression analysis, r² = 0.606, p<0.01). Increasing scores on axis two corresponded to decreasing concentrations of orthophosphate-phosphorus and increasing water temperature (multiple regression analysis, r² = 0.254, p<0.01). Seven regions of the lake ordination were identified by cluster analysis (Fig. 5, Table 7). With the exception of cluster 5, which included only east-side subalpine lakes, the clusters contained mixtures of lakes from several vegetation zones (Table 8). Nonetheless, cluster 3, which was located on the left-side of the graph, included six glacially influenced west-side alpine and subalpine lakes, and cluster 4, which was located on the right-side of the graph, contained two high-forest lakes and four low-forest west- side lakes. Environmental, physical, and chemical characteristics among lakes in the seven clusters are shown in Appendix 1.

¹Two samples each

¹Duplicate samples from four lakes deleted.

Lakes in each cluster differed in terms of the dominant taxa and composition of particular taxonomic groups (Table 9). Lakes in cluster 3 were dominated by Chromulina parvula, and cyanobacteria were low in proportional abundance (<5%). Lakes in cluster 7 were dominated by two unidentified chlorophyte taxa. Lakes in cluster 1 were dominated by Chlorellas p. and Aphanocapsa delicatissima,and lakes in cluster 2 were dominated by A. delicatissima. Lakes in cluster 5 were dominated by Ochromonas sphagnalis and Diogenes sp. (a cyanobacterium). Lakes in cluster 6 were dominated by Diogenes, and lakes in cluster 4 were dominated by Aphanotheca clarthrata. Lakes in cluster 5 and cluster 1 had the lowest and highest species diversity, respectively. Variation in cell densities among the lakes in each cluster was high (Table 10). The average cell densities were highest in lakes in cluster 4 and second highest in lakes in cluster 5. Overall, the average number of taxa per lake was 18.6, with an average of 20.6 taxa in east-side lakes and 17.7 taxa in west-side lakes. Cluster 3 had the lowest average number of taxa at 10.8 per sample.

Several aspects of the results supported our hypothesis that the taxonomic compositions of phytoplankton assemblages in NOCA lakes would correspond with the different physical and chemical conditions that exist among lakes in the four vegetation zones. Based on the analysis of samples collected between 1988 and 1993, decreasing elevation and corresponding shifts in vegetation zone from alpine to low- forest were associated with increased water temperature, pH, alkalinity, conductivity, and concentrations of total nitrogen in NOCA lakes. Alpine lakes were high in total phosphorus and nitrate-nitrogen and low in other water quality variables. These lakes were dominated in density by chrysophyte and chlorophyte taxa, but the highest proportional abundances of Cryptophyta and Bacillariophyta also occurred in these lakes. Cyanobacteria were the dominant taxa in low-forest lakes, which were high in temperature, pH, alkalinity, conductivity, Kjeldahl-nitrogen, ammonia-nitrogen and total phosphorus. Chlorophytes, chrysophytes and cyanobacteria were the dominant taxa in subalpine and high-forest lakes. These lakes were intermediate in temperature, water quality, and concentrations of nutrients relative to alpine and low-forest lakes.

The distributions of phytoplankton taxa in the clusters of the ordination of the 1989 samples were not as consistent with the analysis of the data collected from 1988 through 1993 relative to vegetation zone. Most of the clusters contained lakes from different vegetation zones, especially contiguous zones. In no instance, however, were alpine lakes and low-forest lakes located in the same cluster. Nonetheless, there appears to be two extreme types of phytoplankton assemblages in NOCA lakes. Chrysophyte taxa (Chromulina and Ochromonas) were the dominant taxa, and Cryptophyta were subdominant taxa in the glacially influenced alpine and subalpine lakes in cluster 3. The four low-forest lakes and 2 high-forest lakes in cluster 4 were dominated by Aphanotheca. Lakes in the other clusters exhibited considerable variation relative to the taxonomic composition of the phytoplankton assemblages. However, the effects of seasonal and annual changes of the climate on the phytoplankton assemblages in NOCA lakes remain unresolved because most of the lakes were sampled only once in 1989. Temporal patterns of the phytoplankton assemblages in the NOCA lakes will be addressed in the final report for Phase II of the project.

Phytoplankton assemblages in NOCA lakes were similar in total number of taxa, average number of taxa per sample, and ranges of cell densities relative to phytoplankton studies in lakes in Quebec (Pinel-Alloul et al., 1990), Finland (Arvola, 1986), Olympic National Park (Larson et al., 1991) and Mount Rainier National Park (Larson et al., In Press) (Table 11). Collectively, duration of the ice-free period, geographic location within a study area, lake size, flushing rate, transparency, water temperature, hardness, and nutrients were some of the variables shown to be important factors influencing the taxonomic structures of the phytoplankton assemblages in these other studies. In NOCA lakes, however, vegetation zone and associated water quality and concentrations of nutrients were the apparent factors influencing the taxonomic structures of the phytoplankton assemblages.

Pechlaner (1971) noted that phytoplankton assemblages in alpine and subalpine lakes were dominated by chrysophytes or dinoflagellates. Results from more recent studies have supported this generalization in some cases, but in other studies, the phytoplankton assemblages were dominated by chlorophytes, cyanobacteria, or cryptomonads (Table 12). Results from the NOCA studies supported the conclusion that phytoplankton assemblages in mountain lakes are more complex than originally determined by Pechlaner (1971).