Return to Conservation Biology Research on Communities

Hanson, M.A., J. Allen, D. Buitron, M.G. Butler, T. Call, T. Carlson, N. Hansel-Welch, K. Haws, M. Konsti, D. McEwen, G. Nuechterlein, and K.D. Zimmer. 2006. Lake Christina reclamation: ecosystem consequences of biomanipulation. Final report submitted to the State Wildlife Grants Program. 97 pp. State Wildlife Grants Program final report.


Lake Christina, a 3949-acre shallow lake in Douglas County in west-central Minnesota, is nationally recognized as a critical staging area for migrating canvasbacks, and also is a breeding location for a number of unique nongame bird species. Since the 1950s, the lake has alternated between ecological extremes, sometimes characterized by favorable conditions, and at other times of little use as waterfowl habitat. Sustained high water and dense populations of undesirable fishes are believed to be associated with shifts toward high turbidity and other unfavorable limnological characteristics, along with extreme habitat deterioration for waterfowl and other wildlife. Twice previously, and following obvious trends of habitat deterioration (1965, 1987), the lake was "reclaimed" via chemical removal of fish. Extensive scientific monitoring was conducted in association with the 1987 treatment. Limnological and waterfowl-use data were gathered before and after 1987 to assess the nature and causal mechanisms associated with observed changes. Dramatic improvements in water quality features, extensive development of submerged aquatic plants, and increased fall use by migrating ducks followed the 1987 reclamation (Hanson and Butler 1994a, b). Data gathered before and after treatment contributed to improved understanding of ecology and management potential of shallow lakes in North America. Unfortunately, data gathering efforts at Lake Christina have since dwindled, more or less at the same time as habitat quality and suitability for wildlife has again declined. During 2000-2003, water clarity, distribution of submerged macrophytes, and fall use by migrating ducks all indicated that the lake had again stabilized in a deteriorated condition characterized by poor water quality, a sparse community of submerged macrophytes, and limited suitability for diving ducks and other wildlife species. In attempt to stimulate a limnological shift to more favorable habitat conditions, fish were again removed from Lake Christina using rotenone during October 2003. Here, we summarize responses of fishes, limnological features, and wildlife use during 2004-05, the first two years following the fish removal. Although this is a final project report, manuscripts for publication will be developed from these data and interpretations and synthesis here are subject to change pending further data collection and analysis.

We believe the environmental conditions observed during 2004-05 indicate that Lake Christina has entered a period of transition, and is tending back toward the clear-water state. Our results indicated presence of a persistent fish community during spring 2004, approximately 6 months after the October 2003 rotenone treatment. Data presented here also indicated that recruitment by remnant fish was very strong and that, by 2005, a diverse fish community was again present and included benthivorous, planktivorous, and piscivorous species. Disappointing, but not unexpected, was evidence of rapid recovery by populations of bullheads, carp, and fathead minnows during the 2 years immediately following the rotenone treatment.

However, data gathered during 2004-05 also contain strong signals indicating a move towards more favorable ecological conditions. Concomitant lake-wide trends towards higher water transparency during spring periods, increases in abundance of large-bodied herbivorous zooplankton (Daphnia), and changes in abundance and composition of submerged aquatic plants are consistent with outcomes lake managers had hoped to achieve, and with patterns observed following the 1987 rotenone treatment. We note that one of the most encouraging signals we observed following the 2003 rotenone treatment was the sharp increase in Chara during the 2004, first post-treatment year. Considerable evidence indicates that such increases in Chara often portend major ecological shifts towards a clear-water state in shallow lakes and a similar trend was also observed within a year following the 1987 rotenone treatment at Lake Christina. Finally, we emphasize that even if the over-all lake response is similar to that observed following the 1987 treatment (and triggers a distinct shift to the clear-water state), more dramatic, sustained improvements in water transparency may not be evident until 2006, or even later.

Non-target effects of rotenone in shallow lakes and wetlands may be considerable, but are rarely considered in lake rehabilitation studies. For example, Lake Christina has supported breeding western grebes since the late 1960's and a large population was observed using the lake during 2003. Availability of small prey fishes is considered crucial for successful recruitment of western grebes because adults fly infrequently other than during migration. During 2004, and following the 2003 rotenone treatment, adult western grebes returned to Lake Christina, but quickly abandoned traditional nesting areas and left the lake, presumably due to absence of fishes suitable as prey. By 2005, western grebes returned in large numbers and over 300 nests were identified and monitored. This may indicate that non-target effects of rotenone on some colonial waterbirds should be expected, but are short-term in that breeding waterbird populations return in response to recruitment of young fishes.

Comparison among historical relationships has great potential to help researchers identify signals of transition, thus indicating if and when lake-wide changes are underway. Lake managers have continuing needs to identify limnological signals useful for anticipating periods of rapid change, especially when the lake is entering transition to the turbid-water state. This would facilitate better use of less drastic measures to maintain a clear-water state. For example, since 1999, environmental signs showed evidence that the lake was probably transitioning towards the turbid state. In retrospect, we know that this was true. For example, TP:chl a ratios may be important indicators of the ecological state of Lake Christina, and researchers may benefit from monitoring the trends relative to the 3:1 threshold. Alternatively, based on use of indicator species analyses, concern may be justified when higher than usual counts of Bosmina occur. Additionally, it may be possible to use the importance values of Chara to monitor whether the lake is stable or in transition. If Chara shows sharp lake-wide declines (as it did during the period of 1999-2001, then perhaps the onset of a period of deterioration and a shift to the turbid state may be anticipated.

Full document (3368 KB) This is a PDF file. You will need Adobe Acrobat Reader to download it.