Printer Friendly


Predation on early life stages has the potential to influence anuran population dynamics (Biek and others 2002; McHarry 2017), act as an important evolutionary force (Magnusson and Hero 1991), and link anurans to food webs in breeding ponds (Davenport and Chalcraft 2012). There are a wide variety of known embryo predators, including crayfish (Johnson and others 2003), diving beetles (Magnusson and Hero 1991; Johnson and others 2003), caddisfly larvae (Richter 2000), wasps (Warkentin 2000), fish (Ortubay and others 2006), tadpoles (Magnusson and Hero 1991; Petranka and Thomas 1995), snakes (Warkentin 1995), water molds (Blaustein and others 1994; Gomez-Mestre and others 2006), and leeches (Laurila and others 2002). The majority of organisms smaller than the embryos themselves, however, are generally not recognized as potential predators. Zooplankton, for example, have been observed to have only indirect effects on amphibians through phytoplankton consumption or through a shared predator (Davenport and Chalcraft 2012), and amphibians have been found to prey on amphipods (Cuello and others 2006; Gillespie 2013), but not the other way around.

Freshwater amphipods, including those in the genus Crangonyx, are usually considered detritivorous (Schwartz 1992), but in certain conditions they have been documented to consume a variety of small organisms including zooplankton like Daphnia obtusa, aquatic insect larvae, and other crustaceans such as isopods (Hynes 1954; Schwartz 1992; MacNeil and others 1997). Amphipods have also been found to cannibalize each other and scavenge on dead fish (Hynes 1954; Schwartz 1992; MacNeil and others 1997). Although predation of amphibian embryos and larvae has not been documented, amphipods of several species in the genus Gammarus have been observed preying on larval fish and fish eggs in a few circumstances (MacNeil and others 1997). Given that the partial predatory role of amphipods seems to be highly context-dependent, it is possible that under some circumstances they may prey on amphibian embryos. Here, we report amphipod predation on Northern Red-legged Frog (Rana aurora) embryos.

We conducted field studies on R. aurora egg masses at 2 breeding ponds in Humboldt County, California, during winter of 2016 through spring of 2017. We conducted egg mass surveys weekly throughout the entire breeding season to collect basic phenology and abundance data. During each survey, we marked newly discovered egg masses with numbered pin flags and recorded egg-mass depth below the surface, depth of the pond where the egg mass was found, the Gosner stage(s) of embryos (Gosner 1960) within the egg mass, and the presence of dead embryos.

The 1st site was an ephemeral pond with gradually sloping banks located in the Humboldt Bay National Wildlife Refuge (HBNWR). This site was near the coast at 2-m elevation. It was characterized by an open canopy with abundant cattails (Typha spp.) and other emergent pond vegetation. Other commonly observed aquatic species included Pacific Chorus Frogs (Pseudacris regilla), Rough-skinned Newts (Taricha granulosa), and many aquatic invertebrates or larvae such as snails (Gastropoda), diving beetles (Dytiscidae), giant water bugs (Belostomatidae), mosquito larvae (Culicidae), caddisfly larvae (Trichoptera), dragonfly nymphs (Odonata), and an abundance of amphipods (Crangonyx spp.). At the HBNWR site, new egg masses were observed from 5 November 2016 through 8 February 2017, and egg masses were observed hatching from 12 December 2016 through 9 March 2017. Total water depth at locations where an egg mass was initially observed ranged from 6 to 72 cm, with few egg masses in the deepest part of the pond, which reached approximately 1 m in depth. Two-thirds of egg masses were laid in shallow water 19 to 30 cm deep. During the period when egg masses were developing, water temperatures ranged from 4.5 to 12.4[degrees]C.

The 2nd site was a pond with steep banks surrounding most of its perimeter, located approximately 10 km inland of Big Lagoon (BL), California, on land owned and managed by the Green Diamond Resource Company. This pond typically has water throughout the year, but occasionally dries completely during extended dry periods. The site was at approximately 250-m elevation. It was characterized by a nearly completely closed canopy of Coast Redwood (Sequoia sempervirens) and alder (Alnus spp.), an abundance of woody debris, and sparse emergent vegetation. Other commonly observed aquatic species included Pacific Chorus Frogs, Northwestern Salamanders (Ambystoma gracile), and caddisfly larvae. We observed no amphipods and very few other small aquatic organisms/larvae in samples collected from plankton tows at this site, and did not quantify the abundance of organisms during these plankton tows. At the BL site, new egg masses were observed from 6 January 2017 through 20 March 2017, and egg masses were observed hatching from 9 February 2017 through 10 April 2017. Total water depth at locations where an egg mass was initially observed ranged from 15 to >120 cm, with few egg masses in the deepest part of the pond, which reached approximately 2 to 3 m in depth. Two-thirds of egg masses were laid in shallower waters 20 to 76 cm deep. Water temperatures ranged from 4.1 to 9.9[degrees]C during the period when eggs were developing.

To determine R. aurora hatch rates and predation-related mortality rates prior to hatching, we constructed hatching chambers that were 20 x 15 x 10 cm and consisted of a plastic frame with 3 sides and the bottom covered by fiberglass mesh. The mesh consisted of 2-mm x 1.5-mm rectangles that allowed water flow between the hatching chambers and breeding pond. One chamber in each pair, which we labeled the open treatment, had an open top allowing macro-predators access to the embryos. The other chamber, which we labeled the screened treatment, was covered with a lid made of the same mesh as on the sides and bottom to exclude macro-predator access to the embryos.

We established 15 pairs of hatching chambers at the HBNWR site and 11 pairs at the BL site. Throughout the breeding season, hatching chambers were seeded from egg masses present at each site that were at least 2 wk away from hatching. From each egg mass, we took 2 similarly sized chunks containing 21 to 89 embryos, and placed one chunk in the open hatching chamber and the other in the paired, screened hatching chamber. We left the original egg mass in place. Because we sought to minimize handling of the egg mass and egg mass pieces to be placed into the hatching chambers, we did not always end up with the same number of embryos in each chamber. The number of viable embryos for each treatment in a pair varied by up to 36% (mean difference = 10%). Most embryos in the hatching chambers were in a single clump, but in some cases multiple smaller chunks were used instead of a single clump, as well as a few embryos that were inadvertently separated from the main clump, either individually or as a short strand. Chambers were secured to a wooden stake placed within 2 m of the source egg mass so that total water depth and general environmental conditions would be similar for embryos in the chambers and those in the source egg mass. To allow potential aquatic predators access to embryos, we placed hatching chambers at a minimum of 1 cm to a maximum of 7 cm below the water surface at approximately the same depth as the source egg mass. At the HBNWR site, chambers were set up in water ranging from 18 to 48 cm deep, and at BL chambers were set up in water ranging from 20 to >120 cm deep. At the HBNWR site, the experiment ran from 28 November 2016 to 9 March 2017. At the BL site, the experiment ran from 10 January 2017 to 14 April 2017.

On 5 January 2017 at the HBNWR site, we noticed that many embryos had disappeared from a screened hatching chamber and found numerous amphipods (Crangonyx spp.) in this chamber. We also noticed amphipods present in the open water and crawling on undisturbed egg masses. Most amphipod individuals observed in the pond at this time of year were small enough that they could freely swim through the mesh of the screened chambers. To determine if amphipod predation could be responsible for loss of embryos from screened chambers, we created a 2nd type of hatching chamber impenetrable to amphipods. These chambers were smaller than the original hatching chambers, measuring 15 x 11 x 7 cm. The sides and bottoms of these chambers were solid plastic, and the tops were covered with extremely fine weave polyester Drain Sleeve[R] fabric. The fabric had gaps approximately 0.7 mm in diameter, which allowed for water exchange between the chambers and the pond but did not allow entry of amphipods and larger organisms. We again employed a paired design, placing 10 to 27 embryos from the same egg mass in each chamber of a pair (n = 10 pairs). We placed embryos into water that had been strained through the Drain Sleeve[R] fabric and was free of debris or other visible organisms. In half of the chambers, we added amphipods collected from a 1-min plankton tow in the pond using a plankton net with an opening approximately 15 cm in diameter. We did not quantify amphipod densities, but amphipods added to these chambers appeared to be approximately twice as dense as they were in the open water at the beginning of the experiment, and at similar densities to those in the open water when the last embryos were hatching from egg masses in the pond. We ran this experiment only at the HBNWR site from 12 January 2017 to 21 February 2017, using the same methods for set up as with the open and screened hatching-chamber experiment.

For each treatment type, we checked all hatching chambers weekly and recorded the number of embryos present, the number of visibly inviable embryos, Gosner stage (Gosner 1960) of viable embryos, and any signs of hatching (such as empty egg cases, tadpoles in screened containers). Inviable embryos failed to develop over the course of the experiment, whereas other embryos in chambers from the same source egg mass did. Except in a few cases, inviable embryos never developed beyond Gosner stage 16, and most did not develop beyond Gosner stage 12. Most inviable embryos were further distinguished by color; viable embryos were black, but inviable embryos were typically silvery or white. Because hatched tadpoles could easily swim out of open containers, we based hatch rates, predation rates, and inviability rates on what was present in a hatching chamber the week prior to the 1st signs of hatching. We calculated hatching success as the ratio of viable embryos present the week before hatching to the number of embryos originally placed into the container. We calculated inviability rates as the ratio of inviable embryos to the number of total embryos originally placed in the chamber. We calculated predation rates as the ratio of viable embryos that disappeared from a hatching chamber prior to the week before hatching to the number of viable embryos originally placed into the chamber.

Embryo inviability did not vary much between sites or among treatments. In the open treatment at HBNWR, on average 11% of embryos were inviable, and at BL on average 13% were inviable. The average for inviable embryos in the screened treatments at both sites was 14%. In both the amphipod-exclusion and amphipod-addition treatments at HBNWR, on average 9% of embryos were inviable. The reduction of viable embryos in the hatching chambers varied by site and treatment (Fig. 1). On average, embryos in the open chambers were reduced by approximately a quarter in both study ponds (HBNWR = 27%; BL = 21%); and at the HBNWR site, average embryo reduction in the screened chambers (32%) was similar to the reduction in the open chambers. Conversely, at the BL site, there was no reduction of embryo numbers in the screened hatching chambers. At the HBNWR site, we observed the disappearance of only a single embryo in 1 amphipod-exclusion hatching chamber, whereas the number of embryos in the amphipod-addition chambers was reduced on average by 15%, which was approximately 50% fewer embryos lost than what we observed in the HBNWR open and screened chambers. We attribute the reduction in embryo numbers in the treatments at each pond to predation.

Our results show that amphipods prey upon anuran embryos, as evidenced by the loss of embryos in amphipod-addition hatching chambers without a corresponding loss of embryos in paired amphipod-exclusion chambers. We suspect the disappearance of 1 embryo in an amphipod-exclusion hatching chamber represents a miscount of the number of embryos added to the chamber rather than predation, but we cannot rule out the possibility that an amphipod or other predator was present in the chamber and consumed the egg. The similarity of predation rates in the open and screened hatching chambers at HBNWR suggests that most of the predation we observed in our chambers at this site was due to amphipods and other small organisms that could fit through the fiberglass screen rather than to larger predators such as newts, diving beetles, and water bugs entering the containers through an open top. In contrast, at the BL site, where no amphipods or similarly sized aquatic organisms were observed, no embryos were lost from the screened chambers. The result that predation did occur at BL in the open chambers demonstrates that all observed predation at this site must have been due to larger predators such as salamanders or caddisfly larvae. Thus, although predation rates for open chambers were similar between the two sites, the dominant predators were different.

We caution that the predation rate of embryos by amphipods within our amphipod-addition hatching chambers may not reflect amphipod predation rates in the field for several reasons. First, R. aurora egg masses at our study sites typically contain 400-600 eggs embedded in a thick gelatinous mass (BH and MH, pers. obs.). In contrast to the high proportion of embryos within the interior of most egg masses, the majority of embryos used in our hatching chamber experiments were directly exposed to predation. A 2nd consideration is that amphipods in our hatching chambers did not have access to alternative food resources, which were likely abundant in the open pond. Third, we added amphipods to amphipod-addition chambers at higher densities than those observed in the open pond early in the breeding season but similar to the densities we observed by the time all embryos in the pond had hatched. This could have potentially biased the amphipod predation rates that we measured in our experiments to be higher than those realized in the open pond. A 4th factor is that the water exchange in the hatching chambers we used in our amphipod-addition experiments could have been relatively restricted, given the tight-weave covering of the single chamber opening, which possibly could have reduced dissolved oxygen levels or increased temperatures inside the chambers relative to conditions in the surrounding pond. It is not clear, however, how such differences might have affected amphipod predation rates.

This is the 1st published account of amphipods acting as potential predators on amphibians. Previous accounts of amphipod predation on vertebrates have been restricted to predation on the eggs and larvae of various fish species in marine and freshwater systems (Hunter 1981; MacNeil and others 1997). That amphipods are not among the taxa noted in the published literature as potential amphibian embryo predators despite numerous studies on how amphibians respond to such predators (Sih and Moore 1993; Chivers and others 2001; Laurila and others 2002; Johnson and others 2003; Saenz and others 2003; Gomez-Mestre and Warkentin 2007), suggests that the interaction is relatively rare. The rarity of the interaction may be due at least in part to a scarcity of predatory amphipods in ponds where amphibians breed. Outside of our 2 sites in Humboldt County, we performed plankton tows at 5 other R. aurora breeding ponds in California and Oregon, where we found no amphipods. Although our plankton tows were not designed to comprehensively survey the plankton communities at these sites, the absence of amphipods in our samples does suggest that if amphipods are present at these sites, they occur at substantially lower densities than at HBNWR.

Several factors may have favored this interaction at the HBNWR site and not at the BL site. Potential amphipod predators were rarely observed at HBNWR, whereas at BL, Northwestern Salamanders and caddisfly larvae were abundant throughout the egg development period. Some caddisfly and Ambystoma species are known to consume certain benthic organisms and zooplankton (Hildrew and Townsend 1976; Petranka 1998), and may suppress prey populations through predation. An open canopy along with shallow, warm waters (conditions found at HBNWR) can promote phytoplankton growth (McLaren 1963) in addition to increasing abundance of emergent vegetation and algae, resulting in more food resources for both herbivorous-detritivorous organisms and their predators. In contrast, BL's water temperatures were cold, the pond was heavily shaded, and emergent vegetation and algae were sparse. Cold waters with scarce food resources contribute to suppressed zooplankton abundance (McLaren 1963). Relationships between water temperature and planktonic organisms may appear not to apply directly to freshwater amphipods since they are typically classified as benthic (Schwartz 1992). However, in some habitats they are found to be planktonic (MacNeil and others 1997), and at HBNWR we observed Craugonyx spp. swimming in the water column.

Our findings support the conclusions drawn by MacNeil and others (1997) that amphipods do not fit neatly into the mostly herbivorous functional feeding groups to which they are typically assigned; rather, they can be predators, and their role varies considerably depending on the specific system in question. To complicate matters more, macroinvertebrate and vertebrate predators that consume amphibian embryos in some systems may also consume amphipods, such that predation rates by different predator guilds are not additive. This variability in food web structure complicates the determination of which predator-prey interactions will be important to amphibian population growth or evolution. Figuring out how individual predators affect amphibian survival may be less effective than attempting to look at the entire food web or designing experiments to tease apart relationships between certain subsets of organisms in the environment. As with any ecological interaction, the consequences of amphipod predation on amphibians are likely to be highly context dependent.

Acknowledgements.--This material is based upon work supported by the US Army Corps of Engineers, Humphreys Engineer Center and Support Activity Contracting Office under Contract No. W912HQ-15-C-0051, funded through the Department of Defense Strategic Environmental Research and Development Program (Project RC-2512). We thank the Humboldt Bay National Wildlife Refuge and Green Diamond Resource Company for their cooperation. We are grateful to W Gerth for providing the amphipod identification. The paper benefitted from comments by J Abbott, D Garcelon, K McHarry, M House, E Nelson, and 2 anonymous reviewers.


BIEK R, FUNK WC, MAXELL BA, MILLS LS. 2002. What is missing in amphibian decline research: Insights from ecological sensitivity analysis. Conservation Biology 16:728-734.

BLAUSTEIN AR, HOKIT DG, O'HARA RK, HOLT RA. 1994. Pathogenic fungus contributes to amphibian losses in the Pacific Northwest. Biological Conservation 67:251-254.

CHIVERS DP, KIESECKER JM, MARCO A, DEVITO J, ANDERSON MT, BLAUSTEIN AR. 2001. Predator-induced life history changes in amphibians: Egg predation induces hatching. Oikos 92:135-142.

CUELLO ME, BELLO MT, KUN M, DEVITO J, UBEDA CA. 2006. Feeding habitats and their implications for the conservation of the endangered semiaquatic frog Atelognathus patagonicus (Anura, Neobatrachia) in a northwestern Patagonian pond. Phyllomedusa 5:67-76.

DAVENPORT JM, CHALCRAFT DR. 2012. Evaluating the effects of trophic complexity on a keystone predator by disassembling a partial intraguild predation food web. Journal of Animal Ecology 81:242-250.

GILLESPIE JH. 2013. Application of stable isotope analysis to study temporal changes in foraging ecology in a highly endangered amphibian. PloS ONE 8:e53041. doi:10-1371.

GOMEZ-MESTRE I, WARKENTIN KM. 2007. To hatch and hatch not: Similar selective trade-offs but different responses to egg predators in two closely related, svntopic treefrogs. Oecologia 153:197-206.

GOMIZ-MESTRE I, TOUCHON JC, WARKENTIN KM. 2006. Amphibian embryo and parental defenses and a larval predator reduce egg mortality from water mold. Ecology 87:2570-2581.

GOSNER KL. 1960. A simplified table for staging anuran embryos and larvae with notes on identification. Herpetologica 16:183-190.

HILDREW AG, TOWNSEND CR. 1976. The distribution of two predators and their prey in an iron rich stream. Journal of Animal Ecology 45:41-57.

HUNTER JR. 1981. Feeding ecology and predation of marine fish larvae. In: Lasker R, editor. Marine fish larvae. Morphology, ecology and relation to fisheries. Seattle, WA: University of Washington Press. p 33-77.

HYNES HBN. 1954. The ecology of Gammarus duebeni Lilljeborg and its occurrence in fresh water in western Britain. Journal of Animal Ecology 23:38-84.

JOHNSON JB, SAENZ D, ADAMS CK, CONNER RN. 2003. The influence of predator threat on the timing of a life-history switch point: Predator-induced hatching in the Southern Leopard Frog (Rana sphenocephala). Canadian Journal of Zoology 81:1608-1613.

LAURILA A, PAKKASMAA S, CROCHET PA, MERILA J. 2002. Predator-induced plasticity in early life history and morphology in two anuran amphibians. Oecologia 132:524-530.

MACNEIL C, DICK JTA, ELWOOD RW. 1997. The trophic ecology of freshwater Gammarus spp. (Crustacea: Amphipoda): Problems and perspectives concerning the functional feeding group concept. Biological Reviews 72:349-364.

MAGNUSSON WE, HERO JM. 1991. Predation and the evolution of complex oviposition behaviour in Amazon rainforest frogs. Oecologia 86:310-318.

McHarry K. 2017. Influence of canopy cover and climate on early life-stage vital rates for Northern Red-legged Frogs (Rana aurora), and the implications for population growth rates [thesis]. Arcata, CA: Humboldt State University.

MCLAREN IA. 1963. Effects of temperature on growth of zooplankton, and the adaptive value of vertical migration. Journal of the Fisheries Research Board of Canada 20:685-727.

ORTUBAY S, CUSSAC V, BATTINI M, BARRIGA J, AIGO J, ALONSO M, MACCHI P, REISSIG M, YOSHIOKA J, FOX S. 2006. Is the decline of birds and amphibians in a steppe lake of northern Patagonia a consequence of limnological changes following fish introduction? Aquatic Conservation: Marine and Freshwater Ecosystems 16:93-105.

PETRANKA JW. 1998. Salamanders of the United States and Canada. Washington, DC: Smithsonian Institution Press. 587 p.

PETRANKA JW, THOMAS DAG. 1995. Explosive breeding reduces egg and tadpole cannibalism in the Wood Frog, Rana sylvatica. Animal Behaviour 50:731-739.

RICHTER SC. 2000. Larval caddisfly predation on the eggs and embryos of Rana capito and Rana sphenocephala. Journal of Herpetology 34:590-593.

SAENZ D, JOHNSON JB, ADAMS CK, DAYTON GH. 2003. Accelerated hatching of Southern Leopard Frog (Rana sphenocephala) eggs in response to the presence of a crayfish (Procambarus nigrocinctus) predator. Copeia 2003:646-649.

SCHWARTZ SS. 1992. Benthic predators and zooplanktonic prey: Predation by Crangonyx shoemaker (Crustacea; Amphipoda) on Daphnia obtusa (Crustacea; Cladocera). Hydrobiologia 237:25-30.

SIH A, MOORE RD. 1993. Delayed hatching of salamander eggs in response to enhanced larval predation risk. American Naturalist 142:947-960.

WARKENTIN KM. 1995. Adaptive plasticity in hatching age: A response to predation risk trade-offs. Proceedings of the National Academy of Sciences USA 92:3507-3510.

WARKENTIN KM. 2000. Wasp predation and wasp-induced hatching of Red-eyed Treefrog eggs. Animal Behaviour 60:503-510.

Institute for Wildlife Studies; PO Box 1104; Arcata, CA 95518 USA; Submitted 6 March 2018, accepted 26 December 2018. Corresponding Editor: Michael Parker
COPYRIGHT 2019 Society for Northwestern Vertebrate Biology
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2019 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:GENERAL NOTES
Author:Hudgens, Brian; Harbert, Melissa; Parker, Michael
Publication:Northwestern Naturalist: A Journal of Vertebrate Biology
Geographic Code:1U9CA
Date:Sep 6, 2019

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |