Stocking piscivores to improve fishing and water clarity: a synthesis of the Lake Mendota biomanipulation project

SUMMARYY 1. A total of 2.7 × 10⁶ walleye fingerlings and 1.7 × 10⁵ northern pike fingerlings were stocked during 1987–99 in eutrophic Lake Mendota. The objectives of the biomanipulation were to improve sport fishing and to increase piscivory to levels that would reduce planktivore biomass, increase Daphnia grazing and ultimately reduce algal densities in the lake. The combined biomass of the two piscivore species in the lake increased rapidly from < 1 kg ha^?1 and stabilised at 4–6 kg ha^?1 throughout the evaluation period. 2. Restrictive harvest regulations (i.e. increase in minimum size limit and reduction in bag limit) were implemented in 1988 to protect the stocked piscivores. Further restrictions were added in 1991 and 1996 for walleye and northern pike, respectively. These restrictions were essential because fishing pressure on both species (especially walleye) increased dramatically during biomanipulation. 3. Commencing in 1987 with a massive natural die‐off of cisco and declining yellow perch populations, total planktivore biomass dropped from about 300–600 kg ha^?1 prior to the die‐off and the fish stocking, to about 20–40 kg ha^?1 in subsequent years. These low planktivore biomasses lasted until a resurgence in the perch population in 1999. 4. During the period prior to biomanipulation when cisco were very abundant, the dominant Daphnia species was the smaller‐bodied D. galeata mendotae, which usually reached a biomass maximum in June and then crashed shortly thereafter. Beginning in 1988, the larger‐bodied D. pulicaria dominated, with relatively high biomasses occurring earlier in the spring and lasting well past mid‐summer of many years. 5. In many years dominated by D. pulicaria, Secchi disc readings were greater during the spring and summer months when compared with years dominated by D. galeata mendotae. During the biomanipulation evaluation period, phosphorus (P) levels also changed dramatically thus complicating our analysis. Earlier research on Lake Mendota had shown that Daphnia grazing increased summer Secchi disc readings, but P concentrations linked to agricultural and urban runoff and to climate‐controlled internal mixing processes were also important factors affecting summer readings. 6. The Lake Mendota biomanipulation project has been a success given that high densities of the large‐bodied D. pulicaria have continued to dominate for over a decade, and the diversity of fishing opportunities have improved for walleye, northern pike and, more recently, yellow perch. 7. Massive stocking coupled with very restrictive fishing regulations produced moderate increases in piscivore densities. Larger increases could be realised by more drastic restrictions on sport fishing, but these regulations would be very controversial to anglers. 8. If the lake's food web remains in a favourable biomanipulation state (i.e. high herbivory), further improvements in water clarity are possible with future reductions in P loadings from a recently initiated non‐point pollution abatement programme in the lake's drainage basin.