December 12, 2008

Frog Restoration - the State of the Science

Numerous mountain yellow-legged frog recovery projects have been conducted in the last five years, with more in the planning stages. Virtually all of these projects relied on the removal of nonnative trout and most have met with stunning success. As a result of this work, there is now an abundance of information available on the details of how to remove nonnative trout. But how do we go about selecting sites for restoration? What makes the ideal restoration location?

The mountain yellow-legged frog restoration projects conducted to date have had as their goal increasing the amount of fishless habitat available to an existing frog population. These target frog populations are often small and relegated by the presence of trout to marginal habitats (e.g., shallow ponds). In these projects, site selection was straightforward: choose sites that (1) are in close proximity to existing frog populations, (2) contain high quality frog habitat, and (3) have the appropriate characteristics to allow fish population removal using mechanical means (gill netting, electrofishing). Criterion 1 ensures that mountain yellow-legged frogs will be able to recolonize the restoration site following fish removal. For Criterion 2, high quality habitat is generally characterized as lakes deeper than 3 m (10'), located at elevations below 3600 m (11800'), and surrounded by other suitable habitats including fishless lakes, ponds, marshes, and low-gradient streams (see Knapp et al. 2003 for details). Criterion 3 requires that there are no upstream fish populations and that fish from downstream locations are prevented from moving into the site by natural barriers on the interconnecting streams. In addition,
the chance of successful fish removal is increased by limiting lakes to those of small to moderate size (<20 style="font-style: italic;">Batrachochytrium dendrobatidis - "Bd") into the Sierra Nevada has the potential to considerably complicate restoration site selection. Most restoration projects conducted so far have been in areas in which Bd has not yet arrived. In areas where Bd is present, additional site selection criteria will likely be necessary to maximize the chances of successful frog restoration. In particular, we need to understand the conditions that limit Bd and thereby allow the growth and persistence of frog populations. Bd grows best at water temperatures of approximately 20 degrees C (70 F) and ceases growth above 30 C and below 4 C. In landscapes in which Bd is now ubiquitous (e.g., Yosemite National Park) we are currently studying how temperature regimes affect disease outcomes in frog populations. For example, do temperature effects on Bd growth rates result in increased frog survival in high elevation (i.e., cold) habitats compared to those at low elevations? Is frog survival increased when frogs have access to warm-water habitats such as marshes in which temperatures often exceed 25 C? Answers to such pressing questions are critically needed and we are working feverishly on these studies. Given the ongoing spread of Bd across the Sierra Nevada, the information provided by this research should help guide the selection of sites for future frog restoration projects.

Despite the typical focus of fish removal projects on the mountain yellow-legged frog, we can't lose sight of the fact that trout removal benefits not just frogs but a diverse community of vertebrate and invertebrate taxa. I maintain that having at least some fishless lakes and streams in each watershed across the Sierra Nevada is the most foolproof way to ensure the persistence of these native taxa. Sometimes that will mean removing trout from a lake even if frog recovery is unlikely.

Back to The Mountain Yellow-legged Frog Site.


  1. Hi Roland,

    Given your development of a probability model and your vast survey data for most of the lake basins in the Sierras, have you plugged that data into the model to produce maps showing where frog restoration might be appropriate?

    Of course, i'm assuming this type of approach would at least help some of us understand the basis for decisions.



  2. Hi Russ. Yes, we have used the model to identify restoration sites. The results are certainly useful but not surprising - sites with a high probability of restoration success are those that contain suitable habitat and have mountain yellow-legged frogs nearby.

    The biggest limitation of the model-based approach is that the model does not incorporate the effects of Bd. We're working on this issue but it will take quite a few more years before we have a handle on this.

    So, in places like Yosemite sometimes we've had no choice but to choose restoration sites based simply on the presence of frog populations within 10 km that are persisting despite Bd. We've used these populations as sources of frogs for reintroductions into the restored lakes, but the outcome of these reintroductions remains uncertain - it has worked in some places and not in others.