Biomanipulation impacts on gizzard shad population dynamics, lake water quality, and a recreational fishery. Final Report. St. Johns River Water Management District, Palatka, Florida.

Reversing the effects of eutrophication can be challenging and requires the reduction of external nutrient sources and internal nutrient loading. Omnivorous gizzard shad Dorosoma cepedianum can facilitate nutrient loading from the sediments as a consequence of their foraging activity at the sediment-water interface and subsequent excretion of nutrients in the water column. This feeding activity may contribute considerably to the release of nutrients from the sediments in eutrophic Florida lakes.
Biomanipulation via removal of gizzard shad has been proposed as a management strategy for improving water clarity by reducing internal nutrient loading from the sediments. Preliminary studies at Lake Denham, Florida, suggested that strong biomass reductions of gizzard shad using haul seines may reduce phytoplankton biomass. Recently, biomanipulations have been attempted on several lakes of the Harris Chain of Lakes, Florida using gill nets, but the results of these efforts have yet to be experimentally evaluated. Understanding how fish life history metrics respond to density reductions is critical to understanding the potential impact of biomanipulation on lake food webs. We used a whole-lake gizzard shad reduction experiment,hereafter referred to as a biomanipulation, to 1) assess impacts of a commercial gizzard shad removal on their population dynamics (i.e., recruitment, growth, mortality), 2) measure diet contents of gizzard shad to indicate mode of feeding, 3) explore the potential for gizzard shad removal to influence lake water quality, and 4) evaluate the potential for bycatch impacts on black crappie Pomoxis nigromaculatus fisheries.
We tested the hypothesis that gizzard shad removal at Lake Dora would result in compensatory changes in reproductive rates of the gizzard shad population. We sought to understand the mechanisms for compensatory responses by evaluating changes in growth, reproductive investment, maturation schedules, larval fish densities, juvenile survival, and recruitment of gizzard shad.Lakes Eustis and Harris were used as reference sites with no commercial fishing.
Commercial fishing with gill nets (minimum of 4 inch mesh) occurred in March and April of 2005 and January - March 2006. We collected data on gizzard shad population dynamics at all three lakes from November 2004 to May 2007. The total harvest of gizzard shad from Lake Dora was estimated at 124,989 kg (54 kg/ha) in 2005 and 135,095 kg (58 kg/ha) in 2006. Leslie depletion analysis estimated an exploitation rate on vulnerable-sized fish of 0.61 (95% confidence interval = 0.42 to 0.73) in 2005 and 0.46 (95% confidence interval = 0.30 to 0.63) in 2006.
Total biomass reduction for the gizzard shad population was about 40% from both years of harvest combined. Compensatory responses of individual vital rates were weak following biomanipulation with the exception of length-at-maturity. Gizzard shad at Lake Dora matured at a size 40 mm smaller in 2007 than in 2005, and we observed no changes in size at maturity for Lakes Eustis or Harris. We detected no change in growth, the gonadosomatic index (an index of fecundity), and juvenile survival, but we found a small decrease in average larval fish density after fishing at Lake Dora. Despite small changes in vital rates, we found increased gizzard shad recruitment to age-1 from 2005 to 2007 at Lake Dora, indicating no reduction in gizzard shad recruitment despite substantial decrease in population egg production. Changes in individual vital rates that led to increased recruitment may have been very small or were obscured by sampling variation. Age-1 recruitment estimates were uncertain due to low vulnerability of these small fish to the experimental gill nets. Further sampling in 2008 and 2009 will track these cohorts as they become more vulnerable to the gill nets at age 2. If future samples confirm preliminary conclusions from age-1 recruitment estimates, we would conclude that the population compensated through increased reproduction and maintained constant or possibly increased recruitment despite a 40% total biomass reduction. This finding would have important implications for biomanipulation efforts because compensatory reproduction may dampen biomass reductions by maintaining or increasing the numbers of age-0 gizzard shad, even if the mechanisms for compensation are difficult to detect in field data. Future sampling will refine our conclusions regarding recruitment compensation of the gizzard shad population at Lake Dora.
We used a simulation model to evaluate the relative performance of alternative gizzard shad removal strategies. The 4-inch mesh nets used in the commercial fishery showed dome-shaped vulnerability schedules for gizzard shad, with fish not fully vulnerable to the gear until age 3.5 and vulnerability declining after age 4. Our results show that gill net fisheries for gizzard shad are unlikely to cause large total biomass reductions for gizzard shad (i.e., ? 75% declines) under current gear and fishery configurations. Achieving a 75% reduction in total shad biomass, which is often the target in biomanipulation efforts, could only be achieved by 1) use of smaller mesh sizes, especially 3-inch mesh, 2) very high fishing mortality rates, and 3) fishing every year. We chose a 75% biomass reduction target based on literature reviews of many previous biomanipulation studies, but the degree of reduction required to reduce phytoplankton biomass in Florida lakes is unkown. Our results suggest that long-term total gizzard shad biomass reductions are unlikely to exceed 40-50% at Lake Dora or similar lakes without substantial increases in the fishing mortality (i.e., fishing effort) and decreases in gill net mesh size.
Gizzard shad diets were evaluated using stable isotopes of sulfur and gut content analysis. Gizzard shad ?34S values confirmed the ontogenetic changes in the diet composition reported in literature. During the summer of 2006, gizzard shad ?34S signatures showed clear evidence of an ontogenetic shift from water column to benthic food items. The ?34S values of young gizzard shad were initially high (9-10o/oo) and associated with pelagic modes of feeding, but declined rapidly to values between 0.1o/oo and 2.4o/oo once a TL size of 100-200 mm was reached,suggesting increased importance of benthic feeding. Gut content analysis showed that nearly all gizzard shad stomachs contained evidence of both pelagic and benthic feeding. Gizzard shad in the 100 - 200 mm length class probably derive most of their food from the microflora associated with sediment detritus, whereas larger fish likely spend more time in the water column foraging on zooplankton (copepods and cladocerans), although their foreguts still contained plant and mud detritus. The size relationship with ?34S suggested some size-dependent diet shifts to zooplankton in gizzard shad populations.
Our analysis of zooplankton and water quality data from 2003 to 2007 showed no changes in water quality and macrozooplankton biomass following gizzard shad removal at Lake Dora. The removal resulted in no change in chlorophyll a concentration, Secchi depth, or total phosphorus concentration. There were also clearly no changes in copepod or cladoceran biomass. Macrozooplankton communities may be controlled by a number of other factors including juvenile gizzard shad, threadfin shad D. penetense, invertebrate predators, or density of inedible filamentous algae. We expected gizzard shad removal to reduce water column phosphorus and chlorophyll a based on previous studies evaluating omnivore removals. Our results suggest that either 1) these effects are not likely via gizzard shad removal in Florida lakes, or 2) the biomass reduction was not strong enough to elicit a response in the phytoplankton community or total phosphorus concentrations. Although gizzard shad clearly contribute to internal phosphorus loads in eutrophic lakes, the magnitude of this loading relative to external inputs, sediment fluxes, and wind resuspension is unknown. Our results suggested that these other phosphorus loads substantially exceeded those attributable to the two-year gizzard shad removal at Lake Dora.
Black crappie is the primary sport fish targeted by recreational anglers at Lake Dora, and our results show that the population could be negatively impacted by increases in exploitation resulting from either the recreational fishery or bycatch from the commercial gill net fishery for gizzard shad. The estimated recreational exploitation rate in 2006 was approximately the total sustainable exploitation rate, and increases due to recreational fishing and/or commercial bycatch greatly increase the probability of recruitment overfishing. Resource managers must evaluate policy trade-offs to consider the benefit of the gizzard shad removal and the negative impacts of bycatch mortality on recreational fisheries. Total bycatch estimates in 2006 (January - March) were nearly twice as high as total bycatch estimates in 2005 (March - April). These results suggest that bycatch could be reduced by timing the commercial fishing season to prevent fishing during winter and early spring when black crappie are more abundant in open-water areas where gill netting occurs. Bycatch impacts on black crappie fisheries may be acceptable if the gizzard shad reduction is successful in improving water clarity and increasing aquatic macrophyte abundance. Possible management alternatives are to 1) discontinue the gill net fishery to eradicate bycatch and prevent any harm to the black crappie recreational fishery, or 2) increase commercial effort and gizzard shad exploitation to optimize the success of the biomanipulation.
The results of this study showed that continuing the program at the current level of commercial effort did not optimize either management objective at Lake Dora.
Results of this study show that current commercial fishing gear configurations for gizzard shad reductions are unlikely to achieve large (> 75%) reductions in total gizzard shad biomass. We cannot conclude that biomanipulation is not a viable management tool for restoration of Florida lakes, but our results clearly show that 40% biomass reduction over two years did not significantly influence lake nutrients and zooplankton abundance at Lake Dora. Future biomanipulations targeting water quality improvements should seek to maximize biomass reductions for gizzard shad and should be conducted using control lakes to verify any shifts that occur. Lower mesh sizes and higher commercial fishing effort are recommended, but resource managers should recognize that substantial impacts to black crappie fisheries could occur.