Lake Water Chemistry

Time of ice-out increased from low-forest to alpine lakes. Eastside lakes iced-out earlier than did westside lakes. Epilimnetic temperatures were warmest in low-forest lakes and coolest in alpine lakes. Classification did not order lakes relative to chemical characteristics, although westside low-forest lakes differed significantly from other lake classes and were most productive. Little seasonal and annual variation for most chemical characteristics were found. However, chemical differences did mirror environmental and physical differences among lakes. High phosphorus levels separated glacially influenced lakes. Total Kjeldahl-N and total phosphorus decreased with increasing lake depth. Decreasing lake elevation generally was associated with increased water temperatures, pH, alkalinity, conductivity, and nutrients.

Phytoplankton

1. Phytoplankton assemblages from 64 lakes included 153 taxa. Based on cell density counts for all samples, Aphanocapsa delicatissima had the highest proportional abundance (67.5%). Sixty percent of the taxa occurred in few than 20 of the samples (177).

2. Large seasonal and annual variations in the taxonomic structure of the phytoplankton assemblages were observed.

3. When classified by forest type, alpine lakes had the lowest number of taxa per sample and subalpine and high-forest the highest. Proportional cell densities of chlorophytes and chrysophytes were highest in alpine and subalpine lakes, whereas cyanobacteria dominated high-forest and low-forest lakes.

4. In general, lake elevation, and changes in water quality associated with changes in lake elevation, and concentrations of nutrients were most closely associated with the taxonomic structure of the NOCA phytoplankton assemblages.

Introduced Trout

In NOCA trout densities in lakes with reproducing trout were generally much higher (range 250-724 fish/ha) than in lakes with non-reproducing trout, where trout fry are periodically stocked at low densities (mean 179 fry/ha with mean interval between stocking> 5 years).

Larval Salamanders

General Conclusion

Field and laboratory studies support the view that introduced trout can reduce abundance or even eliminate long-toed salamander larvae (Ambystoma macrodactylum) from fishless high-elevation lakes in NOCA. However, the effects of trout appear to be influenced by the reproductive status of trout and lake chemistry.

Specific Conclusions

1. In fishless lakes, mean larval salamander densities, assessed by snorkeling, were positively related to total Kjeldahl-N (TKN), an indicator of the trophic status of lakes. At higher TKN increased abundance of crustacean zooplankton, an important food resource of larval salamanders, likely contributed to increased larval abundance.

2. Differences in mean larval salamander densities between fishless lakes and lakes with trout was related to TKN concentrations and the reproductive status of trout. Mean larval salamander densities in fishless lakes with TKN <0.045 mg/l were not significantly different from larval densities in lakes with reproducing trout or in lakes with non- reproducing trout. However, in lakes with TKN 0.045 mg/l, mean larval densities were significantly higher in fishless lakes than in lakes with reproducing trout. In lakes with TKN 0.095 mg/L larval densities were significantly higher in fishless lakes than in lakes with non-reproducing trout.

3. In laboratory studies, survival of larval A. macrodactylum (and A. gracile) over a 30-day period was significantly lower in experimental ponds with trout than in ponds without fish, corroborating results from field studies. Mean body size of surviving larvae also was significantly lower in ponds with trout than in fishless ponds.

4. In lakes with fish there was a tendency for a greater proportion of larvae to be hidden in substrate materials on the lake bottom than in fishless lakes, although the difference between fishless lakes and lakes with fish was not statistically significant. Laboratory studies revealed significant differences in substrate utilization of surviving larvae between fishless experimental ponds and ponds with fish.

Crustacean Zooplankton

General Conclusion

Results of field studies are consistent with the view that introduced trout can eliminate or reduce abundance of large diaptomid copepods possibly resulting in increased abundance of smaller herbivorous diaptomids in high-elevation lakes in NOCA. These effects of trout appear to be influenced by trout density, lake depth, and lake chemistry.

Specific Conclusions

1. The abundance of large diaptomid copepods (Diaptomus kenai and D. arcticus) was significantly lower in shallow lakes (maximum depth 10 m) with reproducing fish than in lakes with non-reproducing trout or fishless lakes. There was no significant difference in large diaptomid density between lakes with non-reproducing trout and fishless lakes.

2. The abundance of large diaptomids was significantly lower in shallow lakes with high trout densities than in deep (maximum depth > 10m) lakes with reproducing trout.

3. A small herbivorous diaptomid, D. tyrrelli, was found only in shallow lakes with relatively high concentrations of TKN (>0.05 mg/l) and total phosphorous (>0.007mg/l).

4. For lakes with chemical concentrations suitable for the small copepod, D. tyrrelli density was inversely related to large diaptomid density which could imply interaction of large and small diaptomids. D. tyrrelli often was abundant in shallow lakes with high trout densities where large diaptomids were either absent or low in abundance, whereas in lakes with non-reproducing trout and in fishless lakes where large diaptomids were abundant, the small copepod was usually absent.

Global Conclusions Related to Effects of Introduced Trout On Native Biota

1. Our field studies suggest that introduced trout may have the greatest impact on native biota in shallow lakes (10in maximum depth) where reproducing trout reach high densities. Shallow fishless lakes with higher concentrations of TKN are productive habitats for larval salamanders. When reproducing trout reach high densities in these lakes both larval salamanders and large diaptomid copepods are absent or persist at low abundances and the small herbivorous diaptomid D. tyrrelli may be abundant.

2. Trout appear to have the least impact on native biota in deep lakes (>10 m maximum depth) with non-reproducing trout. These lakes generally have low concentrations of TKN and are not productive habitats for larval salamanders. Since larval densities are low in fishless lakes with low TKN, it is difficult to detect statistically significant differences in larval densities between fishless lakes and lakes with fish. In deep lakes large diaptomid copepods may find refuge from trout predation in deeper water during the day and so escape predation from a visually-oriented predator such as trout.

3. Lakes with non-reproducing trout will be a crucial component of NOCA's high lakes management plan because the lakes are common within NOCA. Furthermore, many anglers prefer to fish in lakes with non-reproducing trout because trout densities are low and fish often reach a large size. Lakes where trout are incapable of reproducing because they lack adequate spawning areas may offer the most options for future management. In these lakes fish densities can be regulated by controlling both stocking densities and the interval between stocking. If deleterious effects are observed, fish can be eliminated in a few years through cessation of stocking. Our field studies suggest that non-reproducing trout have little impact on large diaptomid copepods and only affect larval salamanders in lakes that have high TKN (>0.095 mg/l). However, wildfire during 1994, our last scheduled full field season, limited our opportunity to sample lakes with non-reproducing trout. The sample size of lakes with non-reproducing trout was small (N = 7), most of the lakes were sampled only once, and the lakes that were sampled occupied only a narrow range along the TKN gradient. Thus our conclusions on effects of non-reproducing trout are tentative and further research is needed.