A closer look at our research on multiple predator effects |
|
What are multiple predator effects? |
|
Multiple predator effects (MPEs) are a term given to the impacts of two or more predator species on a single, shared prey species. The consumptive effects of two predator species may be additive, or they may be nonadditive, also known as emergent. When effects are nonadditive, they can either be greater than additive, known as risk enhancement (from the prey's point of view), or less than additive, known as risk reduction (again, from the prey's point of view). Emergent (i.e., nonadditive) MPEs often occur because of behavioral changes in one predator's behavior in the presence of another. This type of mechanism often leads to risk reduction. Alternatively, MPEs can occur when prey have contradictory responses to two predator species and so a response to one predator increases their susceptibility to the other predator. This type of mechanism often leads to risk enhancement. When MPEs are caused by traits or behaviors of the predators or the prey, as they often are, then they are one type of a much larger class of indirect effects known as trait-mediated indirect effects. My research has examined several factors that influence MPEs with the goal of making them more predictive (determining when they should occur and be important). With my research I have also strived to clarify the proper way to explore and document MPEs, both empirically and mathematically. |
|
How does trophic structure influence the redundancy of multiple predators? |
|
Multiple predators may have additive or nonadditive consumptive effects on the shared prey and/or interference effects on each other. For species that have overlapping generations, whether or not effects are additive may depend on the population size structure and whether or not intraguild predation occurs. This study examined first was not only a multiple predator effects study, but also examined species redundancy. Click here to examine this study from a species redundancy perspective. |
|
|
The invasion of two species of crab to New England shores provided an opportunity to examine multiple predator effects by two species that both have overlapping generations, with adult and juveniles present simultaneously. As a result, intraguild predation (predators eating other predators) is common, as larger crabs consume smaller crabs. The figure to the left demonstrates the relative strength of intraguild predation and cannibalism in different combinations of large and small crabs of the two species, Carcinus maenas and Hemigrapsus sanguineus. While this predation among different-sized individuals is prevalent, predation is much less common between similarly sized individuals. Large and small crabs of the two species also share some of the same prey items, such the amphipods that were used in this study. |
I conducted an experiment that examined the multiple predator effects by pairs of of these two species in all possible combinations of large and small individuals of these two species. The figure to the right comes from Griffen and Byers 2006 and shows the interference effect (MPE) and the trophic effect (overall consumption) of different combinations of these two species. Symbols that overlap or are close to each other are redundant. The overall pattern is one of trophic and interference redundancy between the two species when all individuals are small and when all individuals are large. The group that had the strongest MPE (greatest risk reduction) was the combination of large Hemigrapsus with small Carcinus. Overall, when paired crabs differed in size, MPEs became stronger in direct correlation with the intensity of intraguild predation. When crabs were the same size and therefore did not engage in intraguild predation, interference effects were less strong than when intense intraguild predation occurred and were less predictable. This study demonstrates the importance of intraguild predation as a driving mechanism underlying multiple predator effects and highlights that when intraguild predation occurs, the strength of multiple predator effects can be predicted by the intensity of intraguild predation among the different predators. |
 |
Determining the mechanisms of predator interference in different habitats |
|
 |
As described above, predation rates of multiple predator species on shared prey often combine nonadditively due to
interference or facilitation among the predator species. These nonadditive interactions are not constant, but differ with habitat type.
Habitat-specific MPEs are particularly likely when intraguild predation occurs and is the driving mechanism underlying the MPE and when the strength of that intraguild predation differs across habitat types. Griffen and Byers 2006 built from the study described above that showed that intraguild predation was strongest when large Hemigrapsus sanguineus were combined with small juvenile Carcinus maenas. This study first experimentally examined predation by large Hemigrapsus on juvenile Carcinus in two habitat types that are prominent on rocky shores within the Gulf of Maine where these two species overlap. The figure to the left demonstrates that adult Hemigrapsus consume more juvenile Carcinus in rock habitats than in algal habitats. From this we predict that MPEs between large Hemigrapsus and small Carcinus should be stronger in rock than in algal habitats. |
We then quantified the effects of these two predatory crabs foraging together on amphipods in both rock and algal habitats. We also examined the mechanism underlying MPEs in each habitat using a series of treatments with nonlethal predators (having both species present, but with one modified so that it could not consume the shared prey). These nonlethal predator treatments allowed us to tease apart the strength of the different mechanisms underlying the MPE in each habitat type. The figure to the right is from Griffen and Byers 2006 and shows that behavioral effects (reduced foraging by Carcinus in the presence of Hemigrapsus) were the strongest mechanism in each habitat type, followed prey switching (Hemigrapsus consuming Carcinus rather than the shared prey), and density effects (predation that reduces the number of Carcinus present and able to consume shared prey) being the weakest mechanism. |
 |
These results demonstrate several things. First, that trait-mediated indirect effects (behavioral effects + prey switching) was stronger than density-mediated indirect effects in this system. Second, as predicted by the strength of predation by Hemigrapsus on Carcinus in the two habitats, that MPEs were stronger in rock than in algal habitats - in fact, each of the three mechanisms was stronger in rock habitats. By extension, this means that the strength of MPEs can be predicted in different habitats based on knowledge of the strength of intraguild predation. |
|
Detecting multiple predator effects |
|
 |
MPEs have been demonstrated in a variety of systems. Two experimental designs have been commonly used to study MPEs: additive designs in which the treatment with two predator species has double the number of predators as the treatments with a single predator species, and substitutive designs where the total number of predators in single-species and multi-species treatments remains constant. Part A of the figure on the left shows these two experimental designs diagrammatically. However, while these two experimental designs have often been used interchangeably, they actually provide complimentary, but different, information. The additive design indicates whether effects of two predators combine additively. The substitutive design indicates whether nonadditive effects of combining predator species are stronger than nonadditive effects of combining individuals of the same species. |
Confusion has also persisted regarding the appropriate statistical model to use when calculating expectations for consumption when multiple predators are combined. In Griffen 2006 I clarified appropriate experimental and statistical procedures to use when examining MPEs. I derived null models for expected prey survival under both the additive and experimental designs from an original model of multispecies interactions by Billick and Case (1994). (See Griffen 2006 for the appropriate equations to use for calculating expected prey survival under each experimental design.) I then conducted an experiment that used both an additive and substitutive design to investigate MPEs when Carcinus and Hemigrapsus adults foraged together on shared mussel prey at two different prey densities. The figure on the right demonstrates the observed prey survival when the two species were combined and the expected combined predation using both the additive and substitutive designs. Observed prey survival was greater than expected based on the additive experimental design at both prey densities (i.e., risk reduction), while observed prey survival was equal to that expected and less than that expected at low and high prey densities, respectively, based on the substitutive experimental design. |
 |
This study demonstrates that the conclusions drawn from MPE studies can depend on the experimental design used. Specifically, at high prey density an additive experimental design resulted in a conclusion of risk reduction (because of predator interference between the two crab species), while the substitutive design resulted in a conclusion of risk enhancement (because interspecific interference was not as strong as intraspecific interference). I therefore advocate using a combined additive/substitutive experimental design as indicated in Part B of the figure shown above on the left. This design yields more complete information, demonstrating whether nonadditive effects occur when multiple predators are combined, and how strong those effects are relative to intraspecific effects. |
|
Effects of predator density on MPEs |
|
Density is an important factor contributing to the overall
function and impact of species within ecological communities. Density is particularly important in interactions
between species, and its consideration has been instrumental in understanding fundamental ecological processes
that occur between interacting species, including interspecific competition, consumer–resource
interactions, predator functional responses, and resulting trophic cascades. The
importance of predator density when only a single predator
species occurs is well-documented. Interference among
conspecific predators generally increases with predator
density due to higher frequency and intensity of interactions, resulting in lower per capita effects of predators on prey. Griffen and Williamson 2008 examined the combined impacts of two
co-occurring predators across a range of densities in an
effort to determine how predator density affects the way that
consumption by multiple predator species combines. We used a combined additive-substitutive design that included 16 different predator treatments, allowing us to examine MPEs when crabs foraged together at densities of 2, 4, 6, or 8 crabs per experimental enclosure (shown in picture on right). |
 |
 |
The figure to the left is from Griffen and Williamson 2008 and shows that MPEs change with predator density. At low predator density Carcinus and Hemigrapsus mussel consumption was less than additive because of predator interference, but was equal to expected under the substitutive model, indicating that interference between the species was similar in strength to interference among individuals of the same species. At higher predator densities prey consumption was equal to expected based on the additive model, but less than expected based on the substitutive model, indicating that interference between the among individuals of the same species was stronger than interference between the species and that once conspecific interference was accounted for, adding the two species together had no further impact on prey consumption. Given the ubiquity of variation in consumer density across most systems, this study highlights the importance of explicitly accounting for predator density when examining the nonindependent impacts of combining multiple predator species. |
Current work in this area |
|
As Hemigrapsus continues to spread northward in its invasion, the region where Hemigrapsus overlaps with Carcinus is moving northward within the Gulf of Maine. In this area there is a strong latitudinal gradient in water temperature. It therefore remains an open question how the MPEs described above vary with water temperature and therefore change through time as this species interaction migrates northward. In addition to this interesting question, I am currently pursuing several different aspects of multiple predator effects, including both positive and negative interactions. I describe three of these here. First, aggregations of foraging fiddler crabs are common on intertidal sand and mud flats throughout the southeastern United States. These aggregations include dense monospecific and heterospecific aggregations of individuals that display cooperative foraging and herding behavior. I am currently investigating the factors, both biotic and abiotic, that lead to aggregation and herding in single species and multispecies assemblages. Second, previous work with one of my graduate students (Rachel Decker) investigated the role of individual personality in reproductive behavior of the sand fiddler crab Uca pugilator. I am currently building off that work to examine the role of individual personality in aggregation and herding behavior. Third, interspecific interactions that occur along the edge of a species' range are often characterized by steep spatial gradients evenness of the species densities. It is unclear how MPEs and the importance of biodiversity at the consumer trophic level differ as species evenness varies. I am currently investigating this question together with John Griffin at the University of Florida. |
|
