Food Consumption and Growth Rates of Juvenile Black Carp Fed Natural and Prepared Feeds

The black carp Mylopharyngodon piceus is a large cyprinid native to eastern China and Russia from approximately 22°N latitude to 57°N latitude that can reach lengths as great as 2 m and weights as great as 80 kg (Nico et al. 2005). Black carp are specialized feeders and primarily consume mollusks when available (Nico et al. 2005). The preference of black carp for mollusks stimulated interest in propagating and stocking these fish outside their native range to control snails that are intermediate hosts for fish and human parasites. Whether a result of intentional introduction, inadvertent introduction (black carp contaminating a shipment of other fishes), or escapes from aquaculture and research facilities, nonnative black carp are now present in waters throughout the northern hemisphere. Populations are considered established in several Eurasian countries (Nico et al. 2005). In the United States, catches of black carp in the wild have been confirmed in southern Illinois (Horseshoe Lake), Missouri (Mississippi River Lock and Dam 24), and, on multiple occasions, in Louisiana (Mississippi, Atchafalaya, and Red rivers). The primary habitat of the black carp includes river channels and floodplain lakes of temperate, lowland rivers that support an abundance of mollusks (Nico et al. 2005). The strong swimming ability of the carp allows them to navigate swift currents of large rivers that are their preferred spawning habitats (Nico et al. 2005; Schofield et al. 2005, Reference S1). Thus, it appears that black carp would be able to establish populations in many of the United States' rivers.

Black carp collected in Louisiana were suspected to be diploid (Nico et al. 2005), and Jenkins and Thomas (2007) confirmed the presence of diploid black carp in the lower Mississippi and Atchafalaya rivers. Diploidy is the normal genetic condition, and these fish are capable of producing viable gametes. Triploidy, which is accomplished by subjecting fertilized eggs to heat or pressure treatment, results in fish that produce nonviable gametes (Rothmann et al. 1991). Black carp producers have used induced triploidy since 1999 (M. Freeze, Keo Fish Farms, personal communication) to produce fish that can not reproduce if they escape from aquaculture ponds. Thus, collection of diploid black carp from the wild indicates these fish may be progeny of wild-spawned fish or are capable of spawning. Populations of black carp in the wild pose a significant threat to populations of native freshwater mussel and snail species, many of which are imperiled and constitute the most endangered group of aquatic organisms worldwide (Turgeon et al. 1998; Lydeard et al. 2004; Bogan and Strong 2008).

Legislation passed by U.S. Congress in November 2007 added black carp to the list of injurious fishes (Verhey 2007). This action outlawed the interstate transport of black carp without a research permit and eliminated the sale of black carp for use in aquaculture (Verhey 2007). These imposed restrictions should help limit the possibility of future introductions into the wild, but the potential presence of diploid black carp already in the Mississippi, Red, and Atchafalaya rivers necessitates an assessment of the black carp's possible effects on mollusk populations. Although diet studies (Shelton et al. 1995; Venable et al. 2000; Ben-Ami and Heller 2001; Ledford 2003) clearly indicate black carp are capable of consuming native mollusks, there is a need for estimates of measured food consumption and growth rates to forecast the magnitude of the effect of black carp on mollusk populations. We are not aware of studies that have measured food consumption or growth rates of black carp. Thus, the objectives of this research were to estimate food consumption and growth rates for age-1 and age-2 triploid black carp fed natural and prepared feed. We then used estimated food consumption rates to estimate black carp biomass that could adversely affect native mollusk assemblages.

Diploid black carp were not available. We obtained certified triploid black carp from commercial aquaculture operations (Hopper Stephens Hatcheries, Inc., Lonoke, Arkansas and Keo Fish Farms, Inc., Keo, Arkansas) and held them in raceways at the Eastern Unit of the National Warmwater Aquaculture Center on the campus of Mississippi State University, Starkville, Mississippi for 21–28 d before feeding trials began. Raceway temperatures were 20 or 25°C, and fish received prepared feed in excess twice per day; we removed uneaten feed with a sieve 1.5 h after feeding.

We determined daily consumption rate (DCR, 100 · weight of food consumed · mean body weight⁻¹ · d⁻¹, where mean body weight is the mean of initial and final body weight) and specific growth rate (SGR, 100[ln(final weight) − ln(initial weight)]/number of days) for small (77–245 g) and large (466–1,071 g) triploid black carp at water temperatures of 20, 25, and 30°C, temperatures near the optimum temperature for the species based on their native range (Nico et al. 2005) and a previous study (26°C; Ledford and Kelly 2006). Unless otherwise specified, we calculated DCR from weight of food as fed. We measured daily consumption rate and SGR for individual black carp in 113 L aquaria in an environmentally controlled room with a 12 h light and 12 h dark photoperiod. Well water (25–26°C) flowed into each aquarium at 0.3 L/min; dissolved oxygen, monitored once daily, remained higher than 5 mg/L with aeration. Water temperature remained at treatment temperatures of 20, 25, or 30 ± 1°C through adjustment of room temperature and use of chillers and immersion heaters. We cleaned aquaria with a siphon vacuum as needed, usually once every 2 d, to remove organic debris.

Use of pelleted prepared feed simplifies measurement of food and energy consumption, and most black carp readily consumed pellets in the raceways. Small black carp fed on 5 mm, extruded, sinking pellets (Zeigler Feeds, Gardners, Pennsylvania), and large black carp received 8 mm, extruded, sinking pellets (Silvercup Feeds, Murray, Utah). Both formulated feeds contained 40% crude protein and had similar proximate composition. The pellets were relatively uniform in weight (5 mm pellets: mean  =  70 mg, SD  =  7.7 mg, n  =  100; 8 mm pellets: mean  =  407 mg, SD  =  48.8 mg, n  =  100). We intended to use feeding studies to estimate food consumption rate in the wild where food is continuously available; however, providing food continuously and measuring consumption were not possible with pelleted feed because the extruded pellets were stable in water for only 2 h before beginning to disintegrate and leach water-soluble nutrients. Therefore, we conducted preliminary feeding trials to establish feeding frequencies. The assumption was that the feeding frequency that resulted in the greatest DCR would best represent the DCR with continuously available food.

Preliminary feeding trials began by capturing fish from the raceways with dipnets, weighing the fish in water, and randomly placing individual fish into the aquaria. We stocked fish from the 20°C raceway into 20°C aquaria and fish from the 25°C raceway into aquaria at 25°C. We acclimated some of the fish from the 25° raceway to 30°C by raising aquarium temperatures 1°C/d. The individually held fish received pellets in excess of consumption one to four times daily for 5 d. We placed a known number of pellets in a 76-mm-diameter mesh tray on the bottom of each aquarium. We removed trays 1.5 h after each feeding and determined food consumption by the weight of consumed pellets (number of pellets consumed multiplied by mean pellet weight). We measured the DCR for three to five fish of each size group at each temperature and at each feeding frequency. We did not do acclimation to feeding in the aquaria for the preliminary feeding trials; however, all fish appeared to be actively feeding in the raceways, and all fish consumed food during at least one feeding event each day in the aquaria. We analyzed differences in DCR among feeding frequencies for each size group of black carp using analysis of variance (ANOVA, SAS Institute 2009) with α  =  0.10, and we used the feeding frequency that resulted in the greatest DCR in all feeding trials.

The DCR of three to five small (79–150 g) black carp receiving feed one to four times each day did not differ among feeding regimens at 20°C (ANOVA, F_3,10  =  1.18; P  =  0.37), 25°C (ANOVA, F_3,10  =  0.16; P  =  0.92), or 30°C (ANOVA, F_3,10  =  0.24; P  =  0.87). Although DCR was not significantly different among feeding regimens, one feeding each day resulted in the greatest DCR (Figure 1). Therefore, small black carp received feed once each day between 1300 and 1500 hours in all growth and food consumption trials using prepared feed.

The DCR of five large (516–922 g) black carp receiving feed once, twice, and four times each day at each temperature was not different among feeding regimens at 25°C (ANOVA, F_2,12  =  1.28; P  =  0.31) but differed at 20°C (ANOVA, F_2,12  =  3.65; P  =  0.06) and 30°C (ANOVA, F_2,12  =  32.24; P < 0.01). Two feedings each day consistently resulted in the greatest DCR (Figure 2). Therefore, large black carp were offered feed in excess twice each day at 0730–0930 hours and 1300–1500 hours in prepared-feed trials.

Trials to measure DCR and SGR began by capturing individual fish from the raceways with dipnets, weighing in water, and randomly stocking into aquaria such that we could measure DCR and SGR simultaneously for both sizes of black carp at each temperature treatment. We hereafter refer to trials conducted simultaneously as a batch, and each batch included up to 30 trials. We obtained fish for 20°C trials from the raceway maintained at 20°C and fish for 25 and 30°C trials from the raceway maintained at 25°C. For fish in the 30°C trials, we raised the water temperature 1.0–1.5°C/d after transferring the fish to the aquaria. Acclimation to aquarium conditions was a necessary compromise between long periods of time to ensure that fish were actively feeding at a stable rate and short periods of time such that initial weight of the fish accurately represented the actual starting weight. Fish received food in excess in 76-mm-diameter mesh trays for 1.5 h, and we measured food consumption rate the same as in the feeding frequency trials. We considered fish to be acclimated to aquarium conditions when DCR exceeded 2% for two consecutive days. Acclimation was achieved in 3–5 d for all except three individuals that we removed from the trials. Trial duration was 30 d post acclimation for small fish and 40 d post acclimation for large fish, time periods as long as or longer than commonly used in similar studies (e.g., Whitledge et al. 2003; Myrick and Cech 2005). We also evaluated acclimation a posteriori by assessing temporal trends in DCR during the feeding trial visually and using linear regression (PROC GLM; SAS Institute 2009) to detect changes in feeding rate (slope significantly different from 0 or a conspicuous trend) that would suggest insufficient acclimation or some affect of confinement. Note that a small increase in DCR would be the expected condition because most fish were gaining weight and we calculated DCR for the mean weight of the fish during the trial.

We tested differences in DCR and SGR among temperatures separately for small and large black carp using mixed model ANCOVA (PROC MIXED, SAS Institute 2009) with temperature as the independent variable, weight as the covariate, and batch as a random variable. Initial models considered the potential interaction of fish weight and temperature. If the interaction was not significant, we evaluated main effects without the interaction. We considered differences significant at α  =  0.10 due to limited sample size and to increase the likelihood of detecting a difference if one existed. We used least-squares means to compare means when we detected a significant treatment effect. Because fish weight significantly affected DCR in some analyses, we used mean DCR and SGR estimated from the mixed-model ANCOVA equations for a weight central to the distribution of the weight of individual fish fed pellets (150 g for small black carp, 800 g for large black carp) rather than arithmetic means to facilitate comparisons among treatment temperatures.

In a separate batch of trials, small black carp fed on live snails Physa sp. collected from local ponds in six 30-d feeding trials each at 20, 25, and 30°C (18 total trials). We weighed approximately 700 snails and added them to each aquarium at the start of each trial, and we weighed and added additional snails twice daily to replenish the number in each aquarium. Mean live weight of individual snails was 22.9 mg (SD  =  1.1 mg, n  =  3,200). We calculated snail consumption from the weight of snails consumed during the trial, which we determined as the total weight of snails offered minus the weight of snails remaining at the end of the trial. We used the same aquarium stocking methods, acclimation protocol, DCR trend analysis to test for acclimation, and tests for effect of temperature on DCR and SGR as for prepared-feed trials.

We compared the DCR of pellet-fed and snail-fed fish based on energy consumed. We converted food consumed as weight to energy consumed from the energy density (J/g dry weight [dw]) of the pellets and snails, which we measured using a Parr Instrument Company Model 1261 calorimeter (Parr Instrument Company, Moline, Illinois) following procedures in Eggleton and Schramm (2002). We determined energy density of the formulated feed from ten 2–3 g wet weight samples from each pellet size dried prior to calorimetry, and we determined energy density of the snails (whole snail) from four 2–3 g wet weight random samples of snails fed each day during the snail feeding trials that were dried prior to calorimetry.

We tested differences in DCR measured as energy (DCR-E) and SGR between pellet- and snail-fed fish by using a series of ANCOVA models as advocated in Littell et al. (2006), each with fish weight as the covariate and batch as a random effect. The first model included fixed effects for each food and temperature combination (two foods, three temperatures) as well as separate slopes for each combination. Hence the model had six intercepts and six slopes, one for each combination of temperature and food. This model allows one to test for a general linear relationship between the response variable (DCR-E or SGR) and fish weight. That is, the hypothesis tested is that slopes for all combinations of food and temperature are equal to zero. Given a significant relationship indicated by rejecting the hypothesis above, a second model can be fit allowing for different slopes for the six different combinations of food and temperature, which allows for testing slope differences among those combinations. Again, the model includes six intercepts and slopes, but now the hypothesis tested is that the slopes differ among the six combinations of temperature and food. If slopes differ among the six combinations, a subsequent model can then be used to investigate whether slope differences are due to food, temperature, or both food and temperature. We accomplished this by including food-by-weight, temperature-by-weight, and food-by-temperature-by-weight interactions in the model. Removing nonsignificant terms from the model can allow isolation of the source of the different slopes. For DCR-E, the three-factor interaction was not significant, and we removed it from the model. Next a model was fit including both food-by-weight and temperature-by-weight interactions. For this model, we determined that the temperature-by-weight interaction was not significant. Removing the temperature-by-weight interaction term resulted in a model with a significant food-by-weight interaction, which allows for common slopes for each temperature group receiving pellets and different but common slopes for each temperature group receiving snails. Hence the final model has, for the six combinations of temperature and food, three equal slopes for the temperature groups using pellets and three equal slopes for the temperature groups using snails.

Food consumption rates of small and large black carp consuming feed pellets varied, and 27 of 67 fish failed to gain weight (Data S1). Slopes of DCR over time ranged from −0.03 to 0.06 for small fish and −0.05 to 0.05 for large fish. Regression analysis of trends in DCR revealed three small black carp fed pellets had R² > 0.50 and slopes ≥ |0.05|, and we removed these fish from the data set. All other fish either had R² < 0.50 or slopes ≤ 0.03, and none had any change in DCR during the trial, such as a sudden increase or decrease in DCR, that would indicate insufficient acclimation. Thus, we considered these fish acclimated and included them in the assessment of DCR and SGR.

The DCR of small fish differed among temperatures (ANCOVA, F_2,52  =  3.20, P  =  0.05); DCR increased with temperature and differed significantly between 20 and 30°C (Table 1). Daily consumption rate was negatively related to fish weight (ANCOVA, F_2,52  =  25.85, P < 0.01), and the fish weight-temperature interaction was not significant (ANCOVA, F_2,50  =  1.30, P  =  0.28). Similarly, the DCR of large fish increased with temperature (ANCOVA, F_2,31  =  10.68, P < 0.01), and DCR was negatively related to fish weight (ANCOVA, F_2,31  =  23.06, P < 0.01). The fish weight–temperature interaction was not significant (ANCOVA, F_2,29  =  1.97, P  =  0.16).

The SGR of small black carp, although substantially greater at 25°C than at 20 or 30°C, did not differ with temperature (ANCOVA, F_2,52  =  1.95, P  =  0.15; Table 1) and was not linearly related to fish weight (ANCOVA, F_1,52  =  1.41, P  =  0.24), and the fish weight–temperature interaction was not significant (ANCOVA, F_2,50  =  0.82, P  =  0.44). The SGR of large black carp differed with temperature (ANCOVA, F_2,31  =  6.58, P < 0.01) and was greater at 25 and 30°C than at 20°C. The SGR was not significantly related to fish weight (ANCOVA, F_1,31  =  0.30, P  =  0.59), and the fish weight–temperature interaction was not significant (ANCOVA, F_2,29 =  0.91, P  =  0.41).

All 23 small black carp feeding on snails had positive growth (Data S1), and trends in DCR over time revealed no changes that would indicate insufficient acclimation. The DCR did not differ among temperatures (ANCOVA, F_2,18  =  0.08, P  =  0.93; Table 1) and was not linearly related to fish weight (ANCOVA, F_1,18  =  0.80, P  =  0.38); the fish weight–temperature interaction was not significant (ANCOVA, F_2,16  =  0.20, P  =  0.82). Although SGR was substantially greater at 25°C than at 20 or 30°C, SGR did not differ among temperatures (ANCOVA, F_2,18  = 1.96, P  =  0.17) and was not linearly related to fish weight (ANCOVA, F_1,18  =  0.33, P  =  0.57); the fish weight–temperature interaction was not significant (ANCOVA, F_2,16  =  0.54, P  =  0.59).

From energy-density analyses, both 5 mm and 8 mm pellets were 98% dw and ranged from 21.03 kJ/g dw (n  =  10, SD  =  0.06) for 5 mm pellets to 21.87 KJ/g dw for 8 mm pellets (n  =  10, SD  =  0.06). Live snails were 34% dw and 4.21 KJ/g dw (n  =  4, SD  =  0.18). The DCR-E was significantly greater for pellets than snails (ANCOVA, F_1,5 =  22.79; P < 0.01), and the lack of a significant food type–temperature interaction (ANCOVA, F_2,70  =  0.93; P  =  0.40) indicated that temperature did not affect the greater DCR-E with pellets (Table 2). The DCR-E was negatively related to fish weight (ANCOVA, F_1,70  =  15.82; P < 0.01), and the fish weight–food type interaction was significant (ANCOVA, F_1,70  =  13.13; P < 0.01). As a consequence of the significant effect of fish weight, DCR-E of the fish in the two food groups converged with fish size, such that differences in DCR-E between pellets and snails were significant for 100 g (P < 0.01), 150 g (P  =  0.01), and 200 g fish (P  =  0.09) but not for 250 g (P  =  0.71) fish. Despite the substantially greater energy consumption rate of pellet-fed fish, SGR did not differ between pellet-fed and snail-fed small black carp (ANCOVA, F_1,5  =  0.13, P  =  0.73; Table 1).

Black carp readily consumed prepared feed and, based on DCR, appeared to acclimate quickly to individual confinement. Nevertheless, high variation occurred in food consumption and growth, and some fish consumed food but lost weight. Substantial variation in consumption and growth is common in laboratory feeding studies. In studies of other fishes, coefficients of variation in consumption and growth rates range from 5 to 89% (e.g., Karas and Thoresson 1992; Zweifel et al. 1999; Mayfield and Cech 2004). The variation in DCR and SGR for black carp in this study were within the range that other species have shown.

The causes for the variation in DCR and SGR are not apparent. Sufficient acclimation to aquarium confinement did not appear to be a problem, as neither strong nor consistent temporal trends occurred in DCR for individual fish included in the analyses. Trial duration may have contributed to the variation, but the duration of feeding trials in this study was similar to or exceeded that generally considered adequate for estimating food consumption and growth for other fishes (Whitledge et al. 2003; Mayfield and Cech 2004; Hartman and Hayward 2007). However, Hartman and Hayward (2007) noted that an increase in duration of trials with fish in individual confinement can increase variation in growth. Also, the variation in DCR and SGR does not appear to be related to food source or feeding regime because DCR was more variable for black carp eating snails ad libitum than for fish receiving pellets once or twice daily, and the variation in SGR was generally similar for both food sources.

Changes in consumption rate across temperatures can be indicative of optimum temperature for consumption, with optimum temperature being the temperature of maximum DCR and SGR above which these parameters decline (Hartman and Hayward 2007). Ledford and Kelly (2006) found consumption of snails was greatest at 26°C for 10–40 g black carp. In this study SGR peaked at 25°C for small and large black carp that fed on pellets, but consumption continued to increase up to 30°C. It is possible that DCR may have peaked at some temperature between 25 and 30°C (i.e., DCR increased to a maximum at a temperature between 25 and 30°C and then declined to a DCR less than the optimum but greater than DCR at 25°C), and the coarse thermal increments in this study could not detect it. Lacking trials at higher temperatures and evidence for declining DCR, optimum temperature for consumption, although not clearly identified in this study, appears to be greater than 25°C. Refining the optimum temperature for consumption and growth requires future research using smaller temperature increments and higher temperatures.

A different picture on optimum temperature for consumption emerges for small black carp that fed on live snails. Consumption rate, although highly variable, changed little from 20 to 30°C. A reason for the absence of an upward trend in DCR with temperatures ranging from 20 to 30°C, as expected for a fish that is native to temperate climates and as shown by trends in DCR of pellet-fed fish, is not apparent. The snails fed were small (10–70 mg) relative to the size of the black carp (115–245 g). Food consumption may have been mechanically limited, in the sense that the snails were being consumed as fast as they could be processed. Small black carp ate 296 to 674 snails per day; ingestion at this rate could be limited by their mechanical ability to process the snails throughout a 24-h period. Although not statistically significant, the sharp decline in SGR from 25 to 30°C suggests that the fish were unable to consume sufficient energy to meet increasing metabolic needs. The effect of metabolism is supported by a similar decline in SGR from 25 to 30°C despite increasing DCR for small fish fed energy-rich pellets. Larger snails containing more energy per item (although not necessarily higher energy density) may have affected temperature–DCR–SGR relationships differently. Alternatively, snails may not have been an appropriate food for small black carp. Nico et al. (2005), based on a review of Chinese and Russian literature, reported that young (size unknown) black carp consumed aquatic invertebrates in their native range and later switched to mollusks when their pharyngeal teeth had suitably developed. The small black carp fed in these trials had developed pharyngeal teeth as evidenced by shed pharyngeal teeth in the aquaria. Nonmolluscan aquatic invertebrates have higher energy densities than mollusks (Cummins and Wuycheck 1971; Eggleton 2001) and would be a desirable energy source for young fish. However, the similar trend of declining growth rate from 25°C to 30°C for pellet-fed small black carp suggest that energy content was not the principal factor limiting growth, and that optimum temperature for growth is between 25°C and 30°C.

Large black carp fed pellets exhibited a different trend in growth rate from small black carp. For large black carp, the growth rate was approximately the same at 25°C and 30°C. A similar growth rate at 25°C and 30°C indicates an optimum temperature most likely exists between 25°C and 30°C as for small black carp. It is uncertain if large and small black carp have a similar optimum temperature for growth due to the wide temperature increments in this study. There may be some physiological changes in the optimum temperature for growth as black carp grow that were not realized in this study. Additional research across a wider range of temperatures and smaller temperature increments with a wider range of black carp sizes may solidify any change in optimum temperature for growth as black carp age, if one exists.

The consistent inverse trends in DCR with fish size for both small and large black carp (i.e., within size groups) that fed on pellets would predict that DCR should be lower for the large black carp than for small black carp, yet mean DCRs were similar for both size groups. Reasons for this discrepancy are not evident, but the decline in DCR within size groups may be, as suggested for black carp fed small snails, a consequence of mechanical limitation of food consumption. The smaller black carp received smaller pellets than the larger black carp. Ingestion of pellets at a fixed rate could account for similar DCR of large and small black carp but declining DCR within a size class. However, if pellet consumption rate limited food intake, greater total daily consumption would be expected with greater number of feeding events each day, which was not the case. Preliminary feeding trials found that maximum consumption occurred with only one (small black carp) or two (larger black carp) feedings per day. Mechanical limitations and number of feedings do not appear to be main factors leading to discrepancies in DCR among small and large black carp. Feeding studies of a range of sizes of black carp receiving different feed types and feed sizes may be necessary to fully understand black carp size-related trends in DCR.

Variation in food consumption rate and discrepancies in energy consumption rate between snail-fed and pellet-fed black carp preclude precise estimation of food consumption rates. In light of these constraints, this study suggests that energy consumption rates (J·g live weight⁻¹·d⁻¹) of 150 g black carp range from 90.2 to 289.9 at 20°C, 84.8 to 308.4 at 25°C, and 89.4 to 349.5 at 30°C, where the low end of the range is based on the DCR-E for snails and the high end of the range is based on the DCR-E for prepared feed. Energy consumption rates (J·g live weight⁻¹·d⁻¹) of 800 g black carp fed prepared feed were 268.8 at 20°C, 357.0 at 25°C, and 441.0 at 30°C. These energy consumption rates were generally similar to those for small black carp receiving prepared feed, suggesting that the minimum-maximum energy consumption ranges for small black carp may be useful for predicting mollusk consumption for a broader size range of black carp. Our DCR and SGR estimates are for triploid black carp. The extent to which the triploid condition affects DCR and SGR is unknown. Cassani and Caton (1986) found growth rate of diploid individuals of the closely related grass carp Ctenopharyngodon idella was greater than that of triploids, but food consumption rate did not differ between diploid and triploid states.

Even at the lower rates of consumption observed in this study, black carp could have a significant impact on native freshwater mollusk populations in North America. From a review of 84 studies, snail wet biomass ranged from 1 to 100 g/m² among different ecosystems with wet biomass in large streams approximating 10 g/m² (Ruehl and Trexler 2011). Snail energy density in this study was 1,437 J/g wet weight, and Cummins and Wuycheck (1971) proposed 1,806 J/g wet weight. Using an average of 1,621.5 J/g wet weight and a mean biomass of 10 g/m² results in an energy density estimate of 162.2 MJ/ha. Native mussel biomass was 4.1–6.2 g dw/m² across three large reaches (i.e., 2,100–4,700 ha) of the Mississippi River (Newton et al. 2011). Converting mussel dw to energy (18,953.5 J/g dw; Cummins and Wuycheck 1971; Payne and Miller 1989) resulted in pool-wide energy densities of 777.1–1,175.1 MJ/ha.

From the food consumption rates for small black carp consuming snails in this study and assuming 180 d when water temperature was 20–30°C (as occurs in the lower Mississippi River [Schramm et al. 2009]), 1 kg of black carp would consume 15.3–16.2 MJ (range reflects temperatures of 20, 25, and 30°C) of snails or mussels. Feeding at rates similar to energy consumed eating prepared food, 1 kg of black carp would consume 52.2–62.9 MJ of snails or mussels in 180 d. Thus, a black carp biomass of 2.6–10.6 kg/ha would consume the snail biomass in 180 d (these simple calculations do not include growth of prey item or of black carp). In the same way, a black carp biomass of 12.4–50.8 kg/ha would consume a mussel biomass of 4.1 g dry weight/m² (777 MJ/ha) in 180 d (Figure 3). We are not aware of any published biomass estimates for black carp populations. In an assessment of fish predation on zebra mussels Dreissena polymorpha, Eggleton et al. (2004) found mean biomass of native North American fishes expected to eat adult zebra mussels ranged from 0.8 to 3.3 kg/ha. If black carp biomass is similar to other mollusciphagic species, black carp could substantially reduce native snail and mussel populations, even at the lowest measured consumption rate. Whereas black carp could readily consume most freshwater snails, many mussel species grow to large sizes, and some have thick, heavy shells. While we doubt that black carp would select and consume all species and ages of unionid mussels, predation by them could definitely have a negative effect on small cohorts and recruitment of unionid mussel species. Strayer and Malcom (2012) suggested that recruitment failure can occur among freshwater mollusk populations when disturbances are present, leaving relict adult populations that are unable to reproduce successfully. Additionally, Downing et al. (1993) found that a species of freshwater mussel (Elliptio complanata) will experience complete fertilization failure when its population density is reduced to critical levels (<10 adults/m²). Reproductive failure is also common among freshwater snail populations, especially where molluscivores are present (Martin et al. 1992; Bronmark and Weisner 1996). Therefore, it is possible that black carp could reduce freshwater mollusk populations to levels at which they will be unable to sustain themselves, whether through direct reduction from predation or by recruitment failure.

This study illuminates difficulties encountered in achieving precise and consistent estimates of food consumption and growth of black carp in controlled, laboratory conditions when the fish are fed prepared feed and live feed. The variation in estimated food consumption precludes precise specification of important physiological parameters that bioenergetics modeling needs and, therefore, definitive prediction of the rates of mollusk consumption in the wild. Nevertheless, even low estimates of food consumption in this study applied to only a portion of a year, infers that low densities of black carp could quickly reduce and eventually eradicate some native mollusks by predation on all life stages or by consuming smaller individuals and blocking recruitment.

Data S1. Feeding trial raw data for small and large black carp receiving prepared feed at 20, 25, and 30°C and small black carp feeding on snails at 20, 25, and 30°C. The file is in Microsoft Excel format (.xls) and is named FeedingTrialData.xls. Within the file are four data tabs; Small Fish-prepared feed, Large Fish-prepared feed, Small Fish-snails trial 1, Small Fish-snails trial 2. Within each data tab the trial number, temperature, tank number, number of pellets consumed or number of snails consumed, percent body weight of pellets consumed or percent body weight of snails consumed, average of consumption (C), variance of C, total C, start weight, end weight, percent weight gain, and specific growth rate (SGR) are given.

Found at DOI: http://dx.doi.org/10.3996/112012-JFWM-101.S1 (277 KB XLS)

Reference S1. Schofield PJ, Williams JD, Nico LG, Fuller P, Thomas MR. 2005. Foreign nonindigenous carps and minnows (Cyprinidae) in the United States – A guide to their identification, distribution, and biology. U.S. Geological Survey Scientific Investigations Report 2005-5041. Tallahassee, Florida.

Found at DOI: http://dx.doi.org/10.3996/112012-JFWM-101.S2 (20 MB PDF)

The design and conduct of this research and interpretation of results benefited from the input of S. Miranda. T. Bowling, C. Mangum, A. Spencer, and the staff at the Eastern Unit of the National Warmwater Aquaculture Center assisted with data collection. The manuscript benefited from reviews provided by three anonymous reviewers, and P. Allen, P. Hartfield, and T. Newton. M. Fondren, manager of the Eastern Unit of the National Warmwater Aquaculture Center, assisted with aquarium system construction and feeding trials. Funding was provided by the U.S. Fish and Wildlife Service and the U.S. Geological Survey, Mississippi Cooperative Fish and Wildlife Research Unit.

Any use of trade, product or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government.