That’s a good question, and if we are to accurately use dispersal information to make predictions about metapopulation dynamics, fitness must be included. Because dispersal costs are high, reproductive success at a new site must actually exceed that of the original site in order to be advantageous. These costs are probably offset by more fit individuals (larger body size) being the ones that disperse longer distances (there is definitely a relationship between size at metamorphosis and survival). For most species, there is a distribution in how long individuals will wait before dispersal, and some may exhibit staged dispersal consisting of multiple intermediate movements.
Long-distance dispersal, despite the costs, may be advantageous by ensuring avoidance of suboptimal habitat which can act as ecological traps; higher reproduction in nearby, suboptimal sites may still have long-term disadvantages to the population (e.g. poor offspring survival, genetic effects related to inbreeding).

Related, very few species use long-distance acoustic calls to facilitate conspecific dispersal because of the high energy costs of such calls. Those species that do use them tend to have higher dispersal rates overall because they favor early-successional habitat which is less predictable in location.

Thanks for these questions that are provoking me to think a lot! To my knowledge, little is known about the genetic components of dispersal in amphibians. There is a good amount of information available describing dispersal rates and distances for many species, but for any particular species I don’t think there is much genetic guidance for dispersal variation. So, predictions of dispersal within a pool system can be made based on different target species, but I think your question was asking: for a given species, can we preferentially target a genetic subset with known dispersal characteristics to promote the desired amount of dispersal?

What may be a more tractable way to get at that is to determine what environmental settings elicit the desired dispersal patterns, which is what I see more current research doing. Because dispersal is influenced by landscape factors (e.g. fragmentation) and population factors (e.g. density-dependence), designing pool networks that account for these influences may be able to produce dispersal that matches the population management goals. Also, research is trying to differentiate metapopulation structure from a “ponds-as-patches” structure, where pond clusters function as a single independent population. This may be mediated by spatial organization (proximity, pool density), which pool designs can vary in order to manage the population as desired. Identifying genetic traits which could be targeted to achieve desired dispersal rates is a good way to attack the problem from a different angle, but may be equally difficult.

Thanks, good observation about the necessary biological assumptions. First off, we are trying to make inferences about dispersal from post-translocation movements, which may or may not be biologically equivalent to dispersal. Translocations are an effective method because of the ability to standardize motivations for moving, but dispersal itself is a complex behavior which may manifest itself differently under natural conditions. Indeed, some of the next analysis stages will examine how the movement paths relate to other models of movement or dispersal.

Most of the movement paths recorded here have strong directionality (like the truncated example in the figure), suggesting dispersal as opposed to foraging movements often associated with Levy flights/random walks. The step-lengths appear to fit a log-normal distribution (but perhaps a Levy distribution better?). It will be interesting to see if the parameters are distinct for particular portions of the movement path, i.e. is there a marked change in path structure as they perceive, assess, and ultimately select the new habitat, and is this affected by the translocation scale? I also plan to incorporate analyses of tortuosity and fractal dimension of the movement paths to further investigate how the animals are searching the landscape and how they might respond at different scales.

Pretty on-point response, Michael! I’ll keep my eyes peeled for some of your results in the literature. . .

I’m glad you brought up this aspect of the work! Successful wetland restoration will certainly require sound science with tangible applications, dissemination of the knowledge, and willing partners/landowners. Restoration will ultimately hinge on project goals and human values. Luckily, I think recent years have seen more fruitful relationships between on-the-ground managers and agencies striving to offer expert knowledge. For instance, there is a trend toward project prioritizations based on functional significance and public outreach about what ecological services are being provided on people’s land. The topics my research tackles can be incorporated into these planning frameworks.

For this project, my collaborators at SUNY College of Environmental Science and Forestry partnered with a regional network of county soil and water conservation districts, the Upper Susquehanna Coalition. USC implements coordinated project planning amongst the public and agencies to enhance natural resources in the region, such as creating and restoring vernal pools. By partnering with a research university, USC is taking an important step to link the research with the action. I think my colleagues at ESF can speak more to the partnership with USC and what it affords both parties.

Thanks for your interest and compliments, Dr. Anderson! The pool areas in early summer are 39 ± 18 m2 (mean ± SD) and certainly change differently throughout the season. Water depth is continuously recorded and can be used as a covariate in multivariate models of pool selection. I might investigate if and how selection is influenced by the size/depth of an individual’s original pool and/or the pools available during the selection process.

Although I don’t know if I can answer it with my project, it would also be interesting to see if dispersal and habitat selection change during different times of the year in response to changing water depth, or between years based on climate differences. Future work might also experimentally control water depths to simulate dispersal under predicted climate scenarios.

Thanks and let me say that your video was such a great “advanced organizer”. It allowed me to be prepared for and engaged in each aspect of the poster!

Thanks, Dr. Drayton. The question is trying to elucidate the scale of a breeding habitat patch for this species – a single pool or a cluster of multiple pools. As animals assess and select habitat, they do so at various scales (e.g. home range, microhabitat). We want to know if these frogs differentiate between particular pools when they select breeding habitat, or if they select for an entire cluster. You are right that that the frogs can perceive individual pools and recognize them as discrete breeding locations, so the question is whether or not they perceive the cluster of pools as a single patch, and the particular pools within are simply sub-patches that are not themselves selected for. This can be answered by evaluating the pool fidelity of frogs within a cluster. So far, our results suggest there is a lack of fidelity to particular pools and that habitat may be selected at the cluster scale. Additionally, we see that the dispersal characteristics (movement parameters) of the frogs tend to vary based on the scale of movement, suggesting that they are responding to entire clusters differently than to pools within a cluster.

Thanks for your interest, I’d be pleased to discuss more if desired!

Green frogs could be considered habitat generalists, but they still will evaluate breeding habitat based on a number of things, such as vegetative cover, water depth, and yes presence of conspecifics (either attractiveness, or deterrence due to frog density and intra-specific competition). Males tend to arrive at the pools first during the breeding season and stay longer than females. I hope that my data will be able to reveal signs of the evaluation phase by showing changes in the movement path when pool selection occurs. But your questions bring up a good issue that the current study doesn’t test directly: when and why might an individual reject or leave the pool? My work really only looks at part of the entire dispersal process; it leaves out the reasons for emigration from a pool because the translocations control this. A related drawback is that my measure of pool selection does not provide direct evidence of residency and breeding at that pool, only that the frog chose to immigrate there. In fact, some frogs were found to enter a pool and subsequently leave it, evidenced by either the fluorescent trail exiting again right away, or a recapture of the same individual at a different pool later in the season. So, there is a very real question about what constitutes the evaluation phase and how dispersal behaviors relate to the overall habitat selection process.

Maria,

I’ll jump in here…no, R. clamitans is fairly common through much of the eastern US.

I do have a question for Michael. In your study, you translocated frogs between pool clusters at distances over 300m. Did you look at the genetic structure of the frogs before translocating? Is there a chance that you’re breaking up locally adapted subpopulations?

Thanks Maria! R. clamitans (or, recently renamed Lithobates clamitans) is not at all endangered; it is rather ubiquitous throughout its wide geographic range. IUCN lists it as “Least Concern” http://www.iucnredlist.org/details/full/58578/0

This fact makes it conducive to manipulative experiments such as mine. We would hope to be able to extend the inferences made here to similar, threatened species, though that can sometimes be problematic. At any rate, non-endangered amphibian species still provide important ecosystem functions and are frequently at risk of habitat loss and population collapses due to infectious diseases (e.g. ranavirus and chytrid fungus). Additionally, green frogs interact directly with less common amphibian species which are vernal pool specialists, so our understanding of their use of these habitats can be used to manage for the other target species.

Yes, thanks Carole I did not see your post at first… Great question. No, there has been no work done on the genetic structure of the population. It’s possible that there are some genetic differences at that scale, but I would not expect it to reflect localized adaptations. All of these pools are new habitat (constructed in 2010), so the frogs present have colonized from the handful of source populations located throughout. I would suspect that the original colonization included a good amount of genetic mixing to begin with. It’s a very interesting question though that could be investigated within populations across larger scales, or those that have been stable over longer time scales.

Thanks Stacey! I hope to expand on this current work with more research down the line on floodplain wetlands and temporal connectivity.

Thanks Philomena! If only I could have drummed up as much action and enthusiasm as your video…!

At least with the scenes with me in them, I’m running off of nervous energy at trying not to mess up in front of the camera