DISCUSSION

NOCA lakes exhibited a water-quality gradient that gradually changes from low to high concentrations of dissolved solids and total Kjeldahl-N with decreasing lake elevation and increasing water temperature. Phytoplankton cell densities increased along this elevation temperature gradient. These results suggest that phytoplankton productivity tends to increase with decreasing lake elevation and the associated changes in water quality. Furthermore, both species richness and species heterogeneity were positively correlated with the concentration of total phosphorus. The absence of an elevation-temperature gradient for phosphorus gradient in NOCA lakes was due to the relatively high concentrations of phosphorus in lakes in the low forest zone and in alpine lakes and 1 subalpine lake receiving turbid-glacial effluent. When the turbid lakes were removed from the database, total phosphorus increased in concentration with decreasing lake elevation (data not shown). Therefore, it appears that the importance of phosphorus to species richness and species heterogenity was dependent on elevation and temperature in non-glacially-turbid lakes, but for all lakes, both variables were independent of elevation and temperature from a statistical standpoint. These results were similar to those of Nauwerck (1994) who showed that phytoplankton species richness increased with increasing concentration of phosphorus in northern Swedish Lapland lakes.

The increase in relative abundance of cyanobacteria with decreasing lake elevation at NOCA probably corresponds to the associated increases of pH, alkalinity, conductivity, and concentrations of nutrients (Gerloff and Skoog 1957, Fogg et al. 1973, Shapiro 1973, Healey 1982). A similar distributional pattern (based on biomass) was observed by Nauwerck (1994) for cyanobacteria in northern Swedish Lapland lakes. In high-elevation lakes in Mount Rainier National Park (MORA) and Olympic National Park (OLYM), which are located approximately 150 km from NOCA in western Washington, cyanobacteria were low in density (Larson et al. 1994; 1995). However, the elevation range of the lakes sampled at MORA and OLYM was narrow m comparison to NOCA study lakes in that alpine and low-forest lakes were not sampled. In contrast to cyanobacteria, Sandgren (1991) demonstrated that chrysophytes decrease in abundance with increased lake pH, alkalinity, conductivity, and nutrient concentrations. The results of the present study and those of Nauwerck (1994) were consistent with this generalization. At MORA and OLYM, the phytoplankton assemblages of most lakes were dominated by chrysophytes (Larson et al. 1994, 1995). Although chlorophytes can be a common component of the phytoplankton assemblages in oligotrophic montane lakes (Kosswig 1967, Priddle and Happey-Wood 1983, and Nauwerck 1994), at NOCA, chlorophytes were common in alpine, subalpine and high-forest lakes, but were much less common in low-forest lakes. Cryptophytes, diatoms, and dinoflagelates were in low relative abundances in NOCA lakes, a result consistent with those of Rosen (1981), Nauwerck (1994) and Larson et al. (1994, 1995).

Lake (sample) CA ordinations based on algal taxonomic composition indicated that most alpine and subalpine lakes and some lakes in the forest zones had similar phytoplankton floras. However, four subalpine lakes and a group of low-forest lakes and high-forest lakes, located on the far-right side of axis 1 (Figure 3.2), were separated from the other lakes. Although the second axis was not correlated with the environmental variables, increasing scores on the first axis of the CA were correlated with decreasing elevation and increasing temperature, alkalinity, and concentrations of total Kjeldahl-N and ammonia-N. This gradient was practically the same as the environmental gradient associated with phytoplankton cell densities, and provided additional evidence of an increase in lake productivity with decreasing elevation.

The CCA analysis provided evidence that the spatial distributions of some phytoplankton taxa were correlated with either an elevational-alkalinity-total Kjeldahl-N gradient or dissolved nitrogen (nitrate-N or ammonia-N) gradients (Figure 3.4). A chrysophyte, a chlorophyte, a cryptophyte and a cyanobacteria at the positive end of the elevation-alkalinity-total Kjeldahl-N gradient had their highest relative abundances in low-forest lakes (RIDL, WILL, HOZO, THUN, COON, PAN2; Figure 3.3). Five taxa, 3 chrysophytes, a chlorophyte, and a cyanobacteria, were at the positive end of the nitrate-N gradient. These taxa had their highest relative abundances in alpine and subalpine lakes (WILE, MR13, TTAR, TRIL, TRIU; Figure 3.3). The highest relative abundance of one chrysophyte was associated with above average concentrations of ammonia-N in high-forest lakes (NERT, LS1, LS2; Figure 3.3). The other taxa were located near the origin of the two axes, indicating that their associations with the environmental variables were similar. Thus, within the broad context of the alkalinity-total Kjeldahl-N-elevation gradient, these results suggest that nitrogen species were important variables for identifying taxonomic differences of the phytoplankton assemblages among NOCA lakes.

CCA also was used to examine relationships between phytoplankton taxa and the vegetation zones (Figure 3.5). A continuous distribution of taxa was shown from the subalpine zone through the low-forest zone, with a large number of taxa occurring primarily in the subalpine and high-forest zones. Three species (Haematococcus lacustris, Dactylococcopsis acucularis, and Chroomonas acuta) occurred primarily in alpine lakes, whereas five taxa (Cyclotella comita, Gymnodinium sp., Peridinium inconspicuum, Rhodomonas minuta var. nanoplantica, and Chromulina parvula) co-occurred in the alpine, subalpine, and high-forest zones. At the division level, cyanobacteria, chlorophytes, diatoms, and cryptophytes were present in lakes at all vegetation zones. However, Cyclotella stelligera had its highest relative abundances in alpine and subalpine lakes, and therefore its location on the ordination was between the two vegetation zones. A similar argument could be made for the chrysophyte, Chromulina parvula. Overall, chrysophyte taxa had their highest relative abundances in the subalpine lakes and lakes in the forest zones. Dinoflagellate taxa had their highest relative abundances in alpine and subalpine lakes.

Phytoplankton assemblages in NOCA lakes were similar in the average number of taxa per sample and ranges of cell densities to phytoplankton in lakes in OLYM (Larson et al. 1995) and MORA (Larson et al. 1994), but were lower in the total number of taxa (Table 3.7). Duration of the open-water period, geographic location within a study area, lake size, elevation, flushing rate, transparency, water temperature, hardness, and nutrients were some of the variables determined to be important factors influencing the taxonomic structure of phytoplankton assemblages in these and other studies, e.g., Pechlaner (1971), Earle et al. (1986), Earle et al. (1987), Stoddard (1987), Paloheimo and Fulthorpe (1987), Morris and Lewis (1988) and Knoechel and Cambell (1988). In NOCA lakes, elevation and associated changes in water quality and concentrations of nutrients appeared to be the primary factors influencing the taxonomic structure of the phytoplankton assemblages. The role of zooplankton grazing on the densities and taxonomic structures of phytoplankton assemblages in NOCA lakes remains unresolved.

Table 3.7. Total number of phytoplankton taxa, average number of taxa per sample, and the ranges of cell densities in lakes in Olympic National Park, Mount Rainier National Park, and NOCA (1989).

¹55 samples.
²52 samples.