Effects of habitat complexity on intraguild predation and cannibalism in an assemblage of size-structured predators Shannon K. Carter Austin, Texas Baylor University Waco, Texas Class of 2014 Research Experience for Undergraduates Research Project Proposal Mentor: Dr. Patrick Crumrine University of Virginia Blandy Experimental Farm Summer 2013 SUMMARY Predator-prey interactions are among the most widely studied relationships in community ecology, but traditional understanding of energy flow represented in food webs fails to fully explain the complex interactions that take place at the community level. Intraguild predation and cannibalism are two common interactions with substantial effects on community dynamics that are poorly understood and not accounted for in most food web studies. To add even more complexity, both of these interactions are influenced by size-structure and physical habitat composition. Feeding choices of generalist predators are determined more by size of prey than by species identity and densely vegetated habitats provide sites of refuge for prey, thus diminishing the hunting success of these predators. My experiment will examine the effects of habitat complexity on size-structured predator-prey interactions between two size classes of larval aquatic beetles (Cybister fimbriolatus) and larval dragonflies (Anax junius), both of which act as top predators in fishless ponds. These four categories of organisms will be crossed pairwise and resulting predaceous interactions will be measured at high and low habitat complexity. I expect that instances of IGP and cannibalism will decrease with the introduction of habitat complexity. Results of this experiment will further elucidate complex predator-prey interactions in aquatic ecosystems. The project will be shared with the general public at Blandy Experimental Farm and Baylor University. INTRODUCTION AND CONCEPTUAL BACKGROUND Predator-prey interactions in aquatic communities can be very complex because populations are often comprised of individuals that differ in body size. However, food web diagrams often simplify this complexity by ignoring size structure and considering each species as a single entity (Miller and Rudolf 2011). Intraguild predation (IGP) and cannibalism are two common size-structured interactions which influence energy flow through a system, but the impact of these interactions on community structure is underappreciated (Fox 1975a, Polis et al. 1989). Habitat complexity can also vary from one system to another and may influence the strength of these interactions. Theory predicts that IGP should be rare, but IGP is common in aquatic, marine, and terrestrial communities (Polis et al. 1989, Holt and Polis 1997, Crumrine et al. 2008, Arim and Marquet 2004). IGP occurs when two predators competing for the same resource prey on one another (Polis et al. 1989). The interaction can be either symmetrical, signifying that each predator preys on the other (though not necessarily in equal amounts) or asymmetrical, meaning one predator always preys on the other (Polis et al. 1989). Asymmetric IGP is more common in nature; here, the two predaceous species are distinguished as top predator and intermediate predator (Polis et al. 1989). Theoretical models suggest that species engaged in IGP should only coexist when three conditions are met: the intermediate predator is more successful at exploiting the shared resource than the top predator, the top predator acquires great energetic benefits from the intermediate predator, and the shared prey is available at intermediate quantity (Holt & Polis 1997). Despite these limitations, IGP is a common interaction in many communities (Arim and Marquet 2004). In order to explain the commonality of IGP despite restrictions outlined by the mathematical models, it is necessary to consider other contributing variables such as sizestructure within a population. Most IGP interactions result from larger individuals consuming smaller heterospecifics (Crumrine et al. 2008). Wissinger (1992) examined size-structured IGP in an assemblage of larval odonates and found that even as seasonal progression reduces competitive overlap, opportunity for IGP increases because of increasing body size difference between species. More recent studies show that size-structured predator-prey interactions are related to the size range among predators, rather than mean size of these predators (Rudolf 2012). Empirical evidence supporting the hypothesis that greater difference in body size between predators leads to IGP is also available for marine (Armsby & Tisch 2006) and terrestrial (Okuyama 2007) systems. In addition to influencing IGP, size-structure at the population level can promote cannibalism. Cannibalism adds complexity to community interactions and could contribute to coexistence between top predators and intermediate predators more generally than simple mathematical models predict (Rudolf 2007). Cannibalism occurs frequently among aquatic organisms and is most common in sizestructured populations (Fox 1975a). Fox (1975b) found that in the freshwater predator Notonecta hofmanni, cannibalism occurred when different-sized individuals were in close proximity, even if another food source was available. Cannibalistic behavior can be a limiting factor on predator populations that may facilitate the survival of other species (Fox 1975b). Size-selective generalist predators that feed on anything smaller than themselves, regardless of species identity, are among the most likely cannibals in aquatic systems (Yee 2010). Larval odonates are voracious generalist predators that commonly live in resource-limited habitats (Corbet 1999). These factors contribute to the indiscriminate consumption of conspecifics by larval odonates, despite the potential maladaptive implications of cannibalism (Crumrine et al. 2008). Cannibalism in top predators allows top predators and intermediate predators to coexist even in cases where the top predator exploits the shared resource more successfully because it redirects a portion of the top predator’s energy needs away from the intermediate predator (Rudolf 2007). Effects of prey cannibalism on community dynamics, especially related to IGP in size-structured assemblages, have not been adequately explained by empirical data. One study suggests that cannibalistic behavior in the prey can have a significant effect on predator-prey interactions, but the underlying mechanisms are poorly understood (Rudolf 2008). Existing empirical studies have only begun to address the influence that habitat complexity could have on the coexistence of species in IGP systems. A short-term study by Janssen et al. (2007) demonstrated that habitat complexity could promote the persistence of IGP, but size-structure was not considered. Size-structure is important to take into account because it is so influential in the foraging behavior of these voracious predators. Several studies (Yee 2010, Persson and Eklov 1995, Finke and Denno 2006, Swisher et al. 1998) have found that habitat complexity offers refuge for prey and thus may diminish predation rates in an IGP system. Reichstein et al. (2013) suggests that when the intermediate predator was able to take refuge in vegetation, the top predator could not attain adequate nutrition from hunting and died. Additionally, habitat complexity may partially counteract effects of size-structured interactions in communities, creating conditions which may promote IGP (Reichstein et al. 2013). However, robust empirical data addressing the effects of habitat composition on IGP and cannibalism is rather limited (Reichstein et al. 2013). A predator’s behavior influences its response to habitat complexity, and these responses may change the dynamics of community interactions. Hunting strategies of aquatic organisms span a continuum from sit-and-wait to active predators (Crumrine et al. 2008). Sit-and-wait predators take refuge and ambush nearby insects. Habitat complexity could facilitate this foraging strategy, thus increasing the predation rates of sit-and-wait predators (Delclos and Rudolf 2011). Active predators have more opportunities to feed, but also may be more vulnerable to predation themselves (Crumrine et al. 2008). Studies have shown that predation rates of these predators decrease with habitat complexity because their prey take refuge (Reichstein et al. 2013, Persson and Eklov 1995), but Delclos and Rudolf (2011) observed no effect on predation rates. If rates of cannibalism and IGP are affected by feeding strategy, which is highly dependent on habitat complexity, it would logically follow that habitat complexity is an important variable to consider when studying these complex predator-prey interactions. Cybister fimbriolatus and Anax junius are two important predators in fishless aquatic communities. C. fimbriolatus is a highly voracious dytiscid commonly used in predator-prey studies with 3 developmental instars proceeding maturation. C. fimbriolatus is known as an active hunter (Norwood 2012), but can also feed passively, via the sit-and-wait strategy (Yee 2010). A. junius is the common green darner dragonfly with 13 developmental instars. Because they exhibit many instars, A. junius are common subjects in size-structure research (Borror & White 1970). A. junius is predatory but utilizes more passive feeding strategies than C. fimbriolatus, which is a predator of A. junius (Oquendo 2011, Grandinetti 2012). I propose a study to examine the effects of habitat complexity on intraguild predation and cannibalism in an assemblage of size-structured top predators. The top predator of this system will be C. fimbriolatus and the intermediate predator will be A. junius. This will expand on past studies in this system with related objectives and methods. Kawecki (2010) found that prey mortality was directly related to top predator survival and that cannibalistic and IGP interactions increased prey survival. Oquendo’s (2011) study suggests that size structure of the prey is influenced by the type and size of top predator in the system. Finally, Grandinetti (2012) found cannibalism to be much more frequent than IGP among these predators (2012). My proposed study will address the following questions: (1) Does habitat complexity affect the frequency of cannibalism and IGP among A. junius and C. fimbriolatus? (1a) Do A. junius and C. fimbriolatus respond similarly to changes in habitat complexity? (1b) Will the introduction of habitat complexity shift IGP interactions between A. junius and C. fimbriolatus from highly asymmetrical to more symmetrical? In other words, will A. junius become a more effective predator on C. fimbriolatus? (2) Could habitat complexity facilitate coexistence between A. junius and C. fimbriolatus? Hypotheses for the questions raised are as follows: (1) Introduction of habitat complexity will reduce occurrences of cannibalism and IGP because prey will be able to seek refuge. (1a) A. junius’s foraging strategy will be more affected by the introduction of habitat complexity because of its sit-and-wait feeding strategy. C. fimbriolatus is a more active species so the introduction of refuge will not significantly alter its behavior. (1b) It is predicted that A. junius will be a more effective predator on C. fimbriolatus in a more complex habitat because vegetation promotes the foraging strategy of A. junius. (2) While this short-term experiment will reveal some principle effects of habitat complexity on community structure, long-term studies are necessary to prove coexistence. PROJECT DESCRIPTION The experiment will use larvae of the dytiscid water beetle C. fimbriolatus, the green darner dragonfly A. junius, and the blue dasher dragonfly Pachydiplax longipennis. All specimens will be collected from fishless ponds in and around Blandy Experimental Farm. All instars of C. fimbriolatus will be collected from ponds and individuals in first instar will be reared in the lab until they molt and are large enough to use in the experiment. A wide range of instars of A. junius will be collected for use in the experiment. If P. longipennis is not abundant in sampled ponds, another prey species will be selected based on availability. The experiment will consist of a 2 x 2 factorial treatment design with two size classes of C. fimbriolatus (3rd instar and 2nd instar) crossed with two size classes of Anax junius (12th instar and 10th instar). To examine size-structured cannibalism I will also include treatments with the two size classes of each predator resulting is a total of six predator combinations (Table 1). Ten P. longipennis will be present in every treatment as a source of shared prey. Figure 1 illustrates the projected predator-prey interactions between these organisms. Each predator combination will be tested at low and high habitat complexity, creating 12 treatments. Habitat complexity will be manipulated by changing the density of submerged vegetation. Anchored segments of nylon rope will represent vegetation. Habitat complexity treatments will be 52 stems/m2 and 279 stems/m2. These figures are based on the range of abundance of pondweed at shallow depths (Sheldon 1977). To improve statistical power while maintaining manageable workload, each treatment will be replicated two times across three temporal blocks for a total of six replicates per treatment. Fig.1: Arrows indicate predicted predation in the food web. This model represents many predatorprey interactions including intragulid predation and cannibalism. Red arrows illustrate the six groupings used in the experiment. Treatment High Habitat Complexity Low Habitat Complexity Grouping 1 Large C. fimbriolatus Large A. junius Large C. fimbriolatus Large A. junius Grouping 2 Large C. fimbriolatus Small A. junius Large C. fimbriolatus Small A. junius Grouping 3 Large C. fimbriolatus Small C. fimbriolatus Large C. fimbriolatus Small C. fimbriolatus Grouping 4 Large A. junius Small A. junius Large A. junius Small A. junius Grouping 5 Large A. junius Small C. fimbriolatus Large A. junius Small C. fimbriolatus Table 1: Treatments designed to elucidate effects of habitat complexity on intraguild predation and cannibalism in C. fimbriolatus and A. junius. Grouping 6 Small C. fimbriolatus Small A. junius Small C. fimbriolatus Small A. junius The experiment will be conducted in plastic mesocosms (22 cm x 15 cm x 7 cm). All mesocosms will be maintained in the Oak Grove at Blandy Experimental Farm and covered with window screen mesh to prevent wildlife interference and colonization by amphibians and aquatic insects. Mesocosms will be filled with filtered pond water inoculated with phytoplankton and zooplankton. Before being placed in the mesocosm, the instar of each larva will be recorded. As an additional measure of size, headwidth of each A. junius and C. fimbriolatus will be measured with NIH ImageJ. One individual of the appropriate predator type outlined in Table 1 will be added to the mesocosms at the start of the trial. These predator and prey densities reflect natural conditions in fishless ponds (Corbett 1999; Crumrine personal observation). Each trial will last for three days, and will be monitored twice daily. At the end of each trial, surviving specimens will be removed and size measurements will be taken again. During each observation session, predator-prey interactions will be apparent by residual carcasses of consumed prey. C. fimbriolatus and A. junius utilize different feeding strategies which leave prey remains in a condition which reveals the identity of the predator. An empty but complete exoskeleton indicates that C. fimbriolatus was the predator while a torn exoskeleton indicates that A. junius was the predator. Using this knowledge, the predator of each prey will be determined. Prey survival across all predator treatments will be analyzed using 2-factor ANOVA. The proportion of replicates that result in cannibalism or IGP will be determined and analyzed using G-tests. It is hypothesized that large C. fimbriolatus will be the most voracious predator and will consume all other predator and prey present. As indicated by Figure 1, large C. fimbriolatus is predicted to be the top predator in the system. I expect large A. junius will be the second most effective predator, feeding on everything except large C. fimbriolatus In replicates with conspecifics of different sizes, I expect the larger specimen to cannibalize the smaller. In the final grouping, it is hypothesized that the more voracious small C. fimbriolatus will prey on small A. junius. However, this interaction has less experimental evidence and could be more symmetric than hypothesized. These predicted interactions illustrate IGP within the system. IGP will occur between C. fimbriolatus and A. junius, with C. fimbriolatus as the top predator, competing with its prey A. junius for the shared resource. It is also predicted that IGP and cannibalism will be less prevalent as habitat complexity increases because prey will be able to take refuge in vegetation, decreasing the frequency of encounters between predators and prey, and thus decreasing frequency of predaceous interactions. Thirdly, I expect that predaceous interactions by C. fimbriolatus will be less affected by habitat complexity than those by A. junius because C. fimbriolatus is a much more active predator. Lastly, since A. junius’s foraging strategy is promoted by refuge sites, it is predicted that it may be a more successful predator on C. fimbriolatus when habitat complexity is introduced. SIGNIFICANCE AND BROADER IMPACTS This experiment will elucidate a portion of the complex predator-prey interactions that occur within aquatic ecosystems. It will examine the effects of size structure on IGP and cannibalism and demonstrate how these interactions are affected by habitat complexity. The results will show the complexity of aquatic community interactions and the potential vulnerability of these aquatic ecosystems to changes in habitat structure. Knowledge of these interactions can be applied to terrestrial ecosystems and may be particularly useful in developing effective biocontrol strategies in agriculture or to conserve wildlife that may be susceptible to IGP. The research will be performed at the Blandy Experimental Farm in the State Arboretum of Virginia so that the design and results of the experiment can be shared with interested visitors. Results will also be disseminated to K-12th students at a summer camp hosted at Blandy Experimental Farm. Sharing my project with these budding learners will expose them to the scientific process and give them a brief introduction to community interactions. At the end of the summer, results will be shared in a formal presentation at a public conference in Berryville, Virginia. Furthermore, I will bring this project back to Baylor University by presenting it at the Undergraduate Research and Scholarly Achievement Scholar’s Week and sharing it with Baylor’s chapter of BBB. Hopefully, I will also present the project at the annual BBB Biological Honor Society regional conference in Oklahoma. Finally, if significant results are obtained, the research will be published so that it is accessible to the greater scientific community. PROJECTED TIMETABLE June 3rd- June 24th—Find local fishless ponds rich in desired larvae. Most ponds will be within 30 miles of Blandy, but it may be necessary to go farther if larvae are not abundant. Collect study organisms, with back-up specimens in case of death, emergence, or growth outside of desired instar. Obtain all equipment and begin to prepare experimental set-up for trials. June 24th- July 21st—Trials begin in five day increments, repeating trials in three temporal blocks (2 replications per block) through the following 3-5 weeks. A fourth temporal block will be added if time permits. The following outlines daily schedule during data collection weeks: o Monday—collect o Tuesday—collect/set up o Wednesday— begin observation o Thursday—continue observation o Friday—complete observation, tear down July 21st—Data collection complete, begin writing and preparing presentation July 30th – Project Presentation MATERIALS AND BUDGET Item Waders 36 Tablecraft Tote Boxes Nets and Net Bags Clear shoebox containers Nylon Rope Mesh screen (1 100-ft roll) Staple Gun TOTAL Price $172.12 $295.85 $59.51 $27.04 $554.52 Vendor LL Bean Food Service Direct Provided Provided Provided Home Depot Lowe's LITERATURE CITED Arim, M. & Marquet, P.A. (2004). Intraguild predation: a widespread interaction related to species biology. Ecology Letters 7.7: 557-564. Armsby, M. & Tisch, N. (2006). Intraguild predation and cannibalism in a size-structured community of marine amphipods. Journal of Experimental Marine Biology and Ecology 333.2: 286-295. Borror, D.J. & White, R.E. (1970). Peterson Field Guides: Insects. New York, NY: Houghton Mifflin Company. Corbet, P.S. (1999) Dragonflies: Behavior and Ecology of Odonata. Cornell University Press, Ithaca, NY. Crumrine, P.W., Switzer, P.V. & Crowley, P.H. (2008). Structure and dynamics of odonate communities. Chapter 3 in A. Cordoba-Aguilar, ed. Dragonflies: Model organisms for ecological and evolutionary research. Oxford University Press. Delclos, P. & Rudolf, V.H.W. (2012) Effects of size structure and habitat complexity on predator-prey interactions. Ecological Entomology 36.4: 744-750. Finke, D.L. & Denno, R.F. (2006) Spatial refuge from intraguild predation: implications for prey suppression and trophic cascades. Oecologia 149: 265-275. Fox, L.R. (1975) Factors Influencing cannibalism, a mechanism of population limitation in the predator Notonecta hoffmanni. Ecology 56.4: 933-941. Grandinetti, M.E. (2012). Does prey availability affect predator-prey interactions in an assemblage of size structured top predators? (Unpublished Blandy REU Paper). Holt, R.D. & Polis, G.A. (1997). A theoretical framework for intraguild predation. American Naturalist 149.4: 745–764. Kawecki, S (2010). Effects of top predator identity and cannibalism on intraguild predation in larval odonates. (Unpublished Blandy REU Paper). Miller, T.E.X. & Rudolf, V.H.W. (2011). Thinking inside the box: community-level consequences of stage-structured populations. Trends in Ecology and Evolution 26.9: 457-466. Okuyama, T. (2007). Prey of two species of jumping spiders in the field. Applied Entomology and Zoology 42.4: 663-668. Oquendo-Diaz, L.E. (2011). How does body size variation among predators and the type of top predator affect the survival of prey in larval odonate communities? (Unpublished Blandy REU Paper). Persson, L. & Eklov, P. (1995) Prey refuges affecting interactions between piscivorous perch and juvenile perch and a roach. Ecology 76: 70-81 Polis, G.A., Myers, C.A. & Holt, R.D. (1989). The ecology and evolution of intraguild predation: potential competitors that eat each other. Annual Review of Ecology and Systematics, 20: 297-330. Reichstein, B., Schroder, A., Persson, L. & De Roos, A.M. (2013). Habitat complexity does not promote coexistence in a size-structured intraguild predation system. Journal of Animal Ecology 82.1: 55-63. Rudolf, V.H.W. (2007). The interaction of cannibalism and omnivory: consequences for community dynamics. Ecology 88.11: 2697-2705. Rudolf, V.H.W. (2008). The impact of cannibalism in the prey on predator-prey systems. Ecology 89.11: 3116-3127. Rudolf, V.H.W. (2012). Seasonal shifts in predator body size diversity and trophic interactions in size-structured predator-prey systems. Journal of Animal Ecology 81: 524-532. Sheldon, R.B. & Boylen, C.W. (1977). Maximum depth inhabited by aquatic vascular plants. The American Midland Naturalist 97.1: 248-254. Wissinger, S.A. (1992). Niche overlap and the potential for competition and intraguild predation between size-structured populations. Ecology 73.4:1431-1444. Yee, D.A. (2009). Behavior and aquatic plants as factors affecting predation by three species of larval predaceous diving beetles (Coleoptera: Dytiscidae). Hydrobiologia 637.1: 33-43.