Cannibalism of Embryos and Larvae by Adult Woundfin in Intensive Culture: Application to Conservation Propagation

The woundfin Plagopterus argentissimus is a small (generally under 75 mm) omnivorous cyprinid in the Plagopterini tribe (Miller and Hubbs 1960; Figure 1). Native to the lower Colorado River basin, the species probably once ranged from the confluence of the Salt and Verde rivers to the mouth of the Gila River in Arizona, and up the Colorado River into the Virgin River and some of its tributaries in Utah (Miller and Hubbs 1960; Hickman 1987). The species faces serious threats from habitat modification and competition with a nonnative species (red shiner Cyprinella lutrensis; Deacon 1988; U.S. Fish and Wildlife Service [USFWS] 1995). When the species listed as endangered in 1970 (USFWS 1970), woundfin had already realized an 88% range reduction and persisted only in 141 km of the mainstem Virgin River in Utah, Nevada, and Arizona (USFWS 2008). Since its listing as endangered, woundfin have disappeared from at least an additional 56 km of critical habitat in the lower river, and their abundance has declined to precariously low levels elsewhere (USFWS 2008).

The main elements of the woundfin recovery plan are to establish additional populations within the historic woundfin range, maintain two broodstock refugia, and identify sites and protocols for reintroductions (USFWS 1995). Woundfin culture at the Southwestern Native Aquatic Resources and Recovery Center (SNARRC) in New Mexico started in 1987 (USFWS 1995). The stock serves as a genetically diverse refugium population and augments conservation propagation efforts for wild populations (Chen et al. 2011). The facility now operates four culture ponds with 500–1,000 adult fish each, and over the past 5 y, it has produced an average of 10,000 age-0 woundfin annually for conservation stocking efforts in the Virgin River (fish stocked at 4 mo; M. Ulibarri, USFWS, personal communication). Woundfin from SNARRC have served to establish additional refugia populations at Wahweap State Fish Hatchery (Wahweap), Utah and at Bubbling Ponds Native Fish Conservation Facility (Bubbling Ponds), Arizona. In 2010, Wahweap produced 1,055 age-0 fish for stocking (Z. Olsen, Utah Division of Wildlife Resources, personal communication), and Bubbling Ponds produces approximately 75 age-0 woundfin annually (M. O'Neill, Arizona Department of Game and Fish, personal communication). Augmentation efforts have successfully maintained woundfin in the Virgin River (Chen et al. 2011), but the estimated annual number of hatchery-produced woundfin potentially needed for future stocking to facilitate full re-establishment of wild, self-sustaining populations is 100,000–200,000 individuals, or approximately a 10- to 20-fold increase over current production levels (S. Meismer, Virgin River Resource Management and Recovery Program, personal communication). To help meet this potential increase in production, the Bozeman Fish Technology Center (BFTC), Montana, is conducting a multiyear project focused on developing intensive culture techniques for use at the conservation propagation facilities (Webb et al. 2011).

Predation on embryos and larvae by adult woundfin may be an important factor limiting pond culture production. Woundfin are batch spawners (Webb et al. 2009) and are known to prey upon their own offspring (Webb et al. 2011). We experimentally determined the rate of predation on embryos and larvae by adult woundfin under laboratory conditions.

We randomly assigned 10 adult (age-2) woundfin (90 ± 4 mm; mean ± SD) to each of six 0.6-m diameter by 0.3-m-deep experimental tanks at the BFTC for the predation trials. Fish maintenance was at temperatures between 19 and 20°C, with flow-through water and natural photoperiod. Fish acclimated for 2 d before experiments began. We used belt feeders to provide feed (Otohime, C2, Japan) in excess both before and during each trial.

We obtained embryos for experiments by injecting hormones and strip-spawning 2- and 3-y-old fish. We injected males and females with 20 µg/g carp pituitary extract (Argent Laboratories) and then strip-spawned them 24 h postinjection. We reared larvae in 0.6-m-diameter tanks and fed them a mix of Otohime and freeze-dried cyclop-eeze (Argent Laboratories). After all predation trials were complete, we used hand-striping or ultrasound (SonoSite TITAN) to determine the sex of each adult in the predation tanks (Table 1).

We conducted eight embryo predation trials between May 26 and June 21, 2011. Each trial consisted of three randomly selected tanks out of the six tanks (10 adults per tank). In each of these three tanks, we placed a 15- by 23-cm plastic tray containing a layer of glass marbles to imitate spawning substrate. We used noneyed embryos (1–2 d postfertilization) for three trials, and eyed embryos (≥3 d postfertilization) for five trials. During a trial on May 27, the adult fish spawned on the marble substrate, resulting in more than 25 embryos at the end of the experiment. This tank initially held noneyed embryos, and it was not possible to distinguish between the original 25 and the newly fertilized embryos, so we removed this trial from analysis. For subsequent trials, we used either eyed embryos or noneyed embryos that were at least undergoing blastulation to distinguish from newly spawned eggs and fertilized embryos. We randomly placed 25 embryos throughout each tray, except for the trial on May 31 when we used 17 eyed embryos per tray. We left the tanks undisturbed for 4 h before collecting the trays. We then enumerated the remaining embryos and determined their stage of development under a microscope.

We used the same experimental setup and design as the embryo trials to conduct five larval predation trials from June 10 to June 30, 2011. In each trial, we released 15 larvae into three randomly selected tanks of adults. The mesh covering the center of each tank standpipe was 500 µm and precluded larval escapement. Following each trial, we captured the remaining larvae by hand with pipets and hand-nets. At least two people inspected each trial tank to ensure that no individuals, including dead larvae, remained.

We used Wilcoxon rank-sum tests to test for significance of predation on embryos and larvae, as well as to compare predation on noneyed versus eyed embryos. We used a bootstrap with replacement analysis to create 95% confidence intervals for medians. We also used a Kruskal–Wallis test to determine whether there was a relationship between adult sex ratio and predation rate in the experimental tanks. We performed all tests using the statistical program R version 2.11.1 (R Development Core Team 2008) with the exactRankTests package (Hothorn and Hornik 2010).

We detected significant predation effects on both embryos (W  =  210, P < 0.001) and larvae (W  =  45, P  = 0.004). There was a median of three (95% CI, 1–4) embryos eaten per tank (n  =  8 trials; Table 2). We found no significant difference in predation rate on noneyed compared with eyed embryos (W  =  32, P  =  0.07). There was a median of one (95% CI, 0.5–2) larva eaten per tank (n  =  5 trials; Table 3). In both embryo and larval trials, no relationship existed between adult sex ratio and predation rate (P  =  0.829 and 0.276, respectively).

In our experiments, predation on embyros and larvae was substantial. Adults consumed a median of three embryos over 4 h in a tank containing 10 adults, which translates into an average of 0.075 embryos eaten per adult per hour. Using the same calculation, the estimated rate of larval predation was 0.025 larvae eaten per adult per hour.

It is important to note that conditions in the experimental tanks at the BFTC were different from those at the propagation facilities. Most notably, the experimental tanks had very low water turbidity and artificial feed; live food is available in pond culture at the propagation facilities. There is approximately 43 cm of secchi disk visibility at Wahweap (Z. Olsen, Utah Division of Wildlife Resources, personal communication), and there is 46–90 cm of secchi disc visibility at SNARRC (M. Ulibarri, USFWS, personal communication). The water in our experimental tanks had >150 cm of secchi disc visibility. Other studies have found that increased turbidity can significantly decrease rates of predation on larvae by visual predators (e.g., Ohata et al. 2011). If woundfin are locating embryos and larvae visually, then an increase in water turbidity may significantly decrease predation rates. Therefore, the predation rate at the BFTC may represent a maximum or elevated predation rate due to clearer water in a tank environment compared with the propagation facilities.

There are 3,500 adult woundfin in ponds at SNARRC. Assuming a 1∶1 sex ratio, 50% ovulatory success in adult females, 100 eggs per spawning event (Deacon and Minckley 1973), and a single spawning event per female (though female woundfin are batch spawners), a very conservative estimate of the number of progeny from 3,500 adults is 87,500. However, the average annual production at this facility is only 10,000 4-mo-old woundfin for stocking into the river. Even if actual predation rates in the pond culture environment is a small fraction of the maximal rate observed in this study, cannibalistic predation of embryos and larvae could be a substantial contributor to production shortfalls. We recommend that production facilities remove woundfin embryos from the ponds and hatch them in separate ponds. Woundfin in conservation propagation programs currently spawn on gravel-filled baskets placed in ponds. Baskets could be removed daily or several times a week and placed in separate ponds for incubation and rearing. If spawning substrate cannot be removed at least several times a week, personnel could seine the spawning ponds weekly throughout the season to remove larvae. Such seining activity does not appear to cause high rates of mortality or deformity (M. Ulibarri, personal communication).

Please note: The Journal of Fish and Wildlife Management is not responsible for the content or functionality of any supplemental material. Queries should be directed to the corresponding author for the article.

Reference S1. [USFWS] U.S. Fish and Wildlife Service. 1995. Virgin River Fishes Recovery Plan. Salt Lake City, Utah.

Found at DOI: http://dx/doi.org/10.3996/042012-JFWM-029.S1; also available at http://vrhcrp.mesquitenv.gov (4.3 MB PDF).

Reference S2. [USFWS] U.S. Fish and Wildlife Service. 2008. Virgin River fishes, woundfin Plagopterus argentissimus and Virgin River chub Gila seminuda, 5-year review: summary and evaluation. U.S. Fish and Wildlife Service, Utah Field Office, West Valley City, Utah.

Found at DOI: http://dx/doi.org/10.3996/042012-JFWM-029.S2 (558 KB PDF).

Reference S3. Webb MAH, Kappenman K and Fraser C. 2009. Development of Intensive Culture Techniques for Woundfin. Annual Report (2008-2009). U.S. Fish and Wildlife Service. Bozeman, Montana.

Found at DOI: http://dx/doi.org/10.3996/042012-JFWM-029.S3 (4.4 MB PDF).

Reference S4. Webb MAH, Gaylord G, Nistler A, and Fraser C. 2011. Development and optimization of spawning and intensive culture techniques for woundfin, Annual report (2010 - 2011). U.S. Fish and Wildlife Service. Bozeman, Montana.

Found at DOI: http://dx/doi.org/10.3996/042012-JFWM-029.S4 (308 KB PDF).

The authors wish to thank Cal Fraser (Bozeman Fish Technology Center), Manuel Ulibarri (Southwestern Native Aquatic Resources and Recovery Center), Zane Olsen (Wahweap State Fish Hatchery), and Matt O'Neill (Bubbling Ponds Native Fish Conservation Facility) for their expertise in woundfin culture; Luke Holmquist, Sierra Alexander, and Brittany Buchholz for assistance with spawning woundfin; Daniel Drinan for statistical assistance; and Daniel Drinan, Robert Muth and Matt O'Neill for helpful comments on earlier drafts of this manuscript. This manuscript was also improved by the comments from Jan Dean, two anonymous reviewers, and the journal Subject Editor. This study was funded by the Virgin River Recovery Program.

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