Experimental Testing of Buffer Requirements for Amphibians Inhabiting Vernal Pools in Forested

LandscapesKimberly J. Babbitt

University of New Hampshirekbabbitt@unh.edu

Department of Natural Resources and the Environment, Durham, NH 03824

May 31, 2009Demographic responses of amphibians to experimental buffer treatments were variable but body condition of amphibians in the smaller buffers appears to be lower than at sites where no cutting occurred. Cuts surrounding buffers were permeable, but not used, by wood frogs; spotted salamander movement and use of cuts was very dependent on weather. Overall our results suggest that productive forest management andprotection of vernal-pool dependent amphibians can be accomplished but more empirical study on effects of buffer management is needed to assess long term population responses.

Funding support for this project was provided by the Northeastern States Research Cooperative (NSRC), a partnership of Northern Forest states (New Hampshire, Vermont,

Maine, and New York), in coordination with the USDA Forest Service.http://www.nsrcforest.org

Project SummaryForest ecosystem management presents an enormous challenge that requires the use of novel approaches in land stewardship. The central problem is finding the best way to manage forests for timber products, while protecting and maintaining the ecological health of the forest. The protection of vernal pools from timber harvesting is largely dependent on recently developed Best Management Practices (BMPs) that advocate the delineation of terrestrial buffer zones around these habitats to protect amphibians (and other wildlife). However, empirical data on the response of amphibians to buffer management is lacking. To address this question we are examined the effects of terrestrial buffer zone management on amphibian movement and demographics though a landscape-scale field experiment where we manipulated buffer wide by surrounding intact forest with clearcuts. We encircled vernal pools with drift fences and pitfall traps and captured amphibians entering and exiting. We also use radiotracking to track wood frog and spotted salamander movement. Demographic responses were highly variable but body condition of amphibians at wetlands with the smallest (30 m) buffers was lower than from sites with no cutting. Amphibians entered and left wetlands in non-random but variable directions across years and among species suggesting that corridor approaches to management may not be best in this managed landscape. Wood frogs and spotted salamanders moved up to 428 m and 350 m from breeding ponds, respectively. Wood frogs crossed but did not use clearcuts as summer/fall habitat. Salamanders only crossed cuts during a wet year and used mammal burrows extensively, even in clearcuts. There was no evidence that the treatments we examined will result in local population loss, but whether this indicates that a 30-m wide buffer is sufficient is still unclear. However, given the data on body condition it seems unlikely that smaller buffers would be adequate. 100-m wide clearcuts appear permeable to movement and the response of amphibians to buffer treatment are partly due to the width of the cuts. Research on response to wider cuts would be necessary to fully understand the connection between upland habitat and population viability. Overall our results suggest that productive forest management and protection of vernal-pool dependent amphibians can be accomplished but more empirical study on effects of buffer management are needed to assess long term population responses. .

Forest ecosystem management presents an enormous challenge that will require the use of novel approaches in land stewardship. The main problem is finding the best way to manage forests for timber products, while protecting and maintaining the ecological function of forest ecosystems.

Increasing focus has been placed on the protection of vernal pools in recognition of their importance in forested landscapes. Vernal pools are small, shallow ponds characterized by a lack of fish and annual or semi-annual periods of dryness. These habitats are numerically abundant in many forested landscapes in the U.S. and are critical to the survival of many invertebrate and vertebrate species. Vernal pools serve as exclusive or facultative breeding habitat for several species of amphibian, contributing significantly to local and regional amphibian species richness.

Habitat loss and fragmentation are among the most significant threats to amphibian populations. Although vernal pools are critical breeding habitat for several amphibian species, the species that depend on these wetlands spend a majority of their lives in the surrounding terrestrial habitat. Thus, exclusion of surrounding terrestrial areas from protection under wetland statutes and land use planning has serious consequences for amphibians that breed in vernal pools.

Most recently, BMPs have been developed for vernal pool management and the guidelines are largely driven by the needs of amphibians that depend on these sites for critical breeding habitat. As a majority of amphibians are forest-dependent, developing management approaches that optimize resource use while maintaining ecosystem function for amphibians is a critical component of sustainable forestry.

Because amphibians that breed in vernal pools spend a majority if their lives in upland habitat, delineation of buffer zones around vernal pools to protect these habitats and dependent amphibians (and other organisms) from logging impacts has been recommended.

However, little empirical data exist to provide solid, defensible, guidelines about how wide these buffers need to be to protect amphibians populations.

Thus, the purpose of this project was to determine experimentally the utility of forested buffers as part of sustainable forest management approach for vernal-pool dependent amphibians by conducting a field-base, landscape-level experiment to examine the response of vernal-pool organisms to buffers of different sizes.

Major Objectives

Examine demographic responses of amphibians to buffer treatments by tracking population structure over several years.

Examine effects of buffer treatments on amphibian movement and habitat use via radiotracking.

Examine immigration and emigration patterns of adults and metamorphs.

Examine effects of buffer treatment on aquatic macroinvertebrates assemblages, with particular focus on insects that are predators on amphibian larva.

Project Location

Eastern Maine - Hancock and Washington counties Land owned by International Paper/Sustainable Forest

Technologies

Clearcuts done by IP/SFT between September 2003 and March 2004

study sites

100 m

30 m Buffer

100 m

100 m Buffer

Vernal pool

ForestClear cut

100 m

30 m Buffer

100 m

100 m Buffer

100 m

30 m Buffer

100 m

30 m Buffer

100 m

100 m Buffer

100 m

100 m Buffer

Vernal pool

ForestClear cutVernal pool

ForestClear cutVernal pool

ForestClear cut

8 treatment wetlands:30 m (n=4) or 100 m buffer (n=4) surrounded by100 m-wide clearcut

3 reference wetlands: no clearcut

Experimental Design

Clearcut

ForestedBuffer andVernal Pool

Aerial View of a Study Site

Data collection All amphibians entering and leaving wetlands were

identified, enumerated, aged (adult, subadult, metamorph), sexed, weighed and measured (SUL).

Checking pitfall traps, making measurements, example of data sheet.

Data collectionWood frogs (Lithobates sylvaticus) and spotted

salamanders (Ambystoma maculatum) were the two target organisms of most interest thus additional data were collected on those species. Egg mass counts were conducted, individuals were

given a wetland id mark to determine recapture rate, orientation of ingress and egress from the wetlands was examined, and someindividuals were radio tracked.

Demographic Responses -Spotted Salamanders

There was significant variation among wetlands in breeding numbers; variation was less pronounced at reference wetlands

Fewer salamanders bred in 2007 than in 2004 (the reference year)

This drop in #s of salamanders in 2007 was especially pronounced at the 100 m treatment.

Demographic Responses – Spotted Salamanders

There was significant variation in recapture numbers among wetlands.

The # of recaptured salamanders in each year differed more widely at the 30 m treatment than at the reference treatment.

Demographic Responses – Spotted Salamanders

Spotted salamander females at the 30 m treatment weighed less than those at the reference treatment.

There was less of a difference in mass between females and males at the 30 m treatment than at the reference treatment.

Demographic Responses – Spotted Salamanders

Spotted salamander females at the 30 m treatment had worse body condition (mass relative to length) than those at the reference treatment.

There was less of a difference in body condition between females and males at the 30 m treatment than at the reference treatment. Error Bars: +/- 2 SE

Demographic Responses – Spotted Salamanders

Recaptured salamanders at the 30 m treatment had worse body condition than recaptured salamanders at the reference treatment.

There was less of a difference in body condition between the sexes among recaptured salamanders at the 30 m, compared to the reference treatment.

Demographic Responses – Wood Frogs

The number of breeding wood frogs did not differ greatly among treatments but was slightly higher at reference sites.

There was a high degree of variability in breeding numbers among wetlands.

Demographic Responses – Wood Frogs The number of recaptures

was lower and less variable at the 30 m treatment.

Recaptures varied among years and was similar across treatment in 2006.

Overall, there was a high degree of variability in recapture at wetlands, particularly in the reference and 100 m treatments.

Demographic Responses – Wood Frogs Wood frog females at the 30

m treatment weighed less than those at the reference treatment.

There was less of a difference in body condition between females and males at the 30 m treatment than at the reference treatment.

Demographic Responses – Wood Frogs

Wood frog females at the 30 m treatment had worse body condition (mass relative to length) than those at the reference treatment.

There was less of a difference in body condition between females and males at the 30 m treatment than at the reference treatment.Error Bars: +/- 2 SE

Demographic Responses – Wood Frogs Recaptured female wood

frogs at the 30 m treatment had worse body condition than recaptured females at the reference treatment.

There was a less of a difference in body condition between the sexes at the 30 m treatment than at the reference treatment.

Movement/Radiotracking Results We tracked 33 wood frogs in two years. Wood frogs traveled up to 428 m from their breeding ponds. Frogs at 30-m buffer sites consistently traveled through the

buffer, crossed the clearcut and entered the adjacent forest. Frogs at 100-m buffer

sites were more likely to stay in the buffer than cross the cut.

Wood frogs spent little in the clearcut.

Females moved farther than males.

Movement/Radiotracking Results We tracked 40 spotted salamanders in two years. Spotted salamanders traveled up to 350 m from their

breeding pond. Salamander migrations was

significantly affected by weather: in a dry year no salamander crossed the clearcuts but in a wet year nearly half crossed the cuts.

Salamanders spend a majority of time underground in small mammal burrows, including in clearcuts.

##

#

#############

#

######

###

#

5/15

11/6

5/17 - 10/8

5/2 - 7/26

5/31 - 9/165/2 - 5/209/30 - 10/6

5/10 - 5/205/2 - 5/6 4/30

5/31 - 7/7

100 0 100 200 Meters

Clear cut

Buffer

Woods beyond clear cut

Depiction of four patterns of migration behavior at a 100-m-buffer wetland. Different colors are different individuals. One individual moved through the buffer and clearcut into woods beyond the clearcut (red). Another lingered near the clearcut edge, spent the summer in the clearcut, and returned to the edge during the fall (blue). A third stayed in the buffer for the entire season (purple). A fourth individual quickly moved through buffer, and spent the restof the time it was tracked in the clear cut (pinkish-brown).

Did wood frogs and spotted salamanders move in the same directions at a given wetland? No, different directions for 81% of comparisons.

Ambystoma maculatumLithobates sylvaticus

Orientation Results

Note: For all orientation figures data are composites or examples to demonstrate results. Analyses were conducted on 2 species, 2 age classes, immigrating and emigrating animals, across 3 years at each wetland.

Were movement patterns consistent across years? No, movement was in different directions in 91% of comparisons.

2004 2005 2006

Ambystoma maculatum

among years, DF=14, p<0.0001

Orientation Results

2005

2006

Lithobates sylvaticus

among years, DF=14, p<0.0001

2004

Orientation Results

Orientation Results

No, orientation direction was nonrandom in 76% cases for spotted salamanders and 91% of cases for wood frogs.

So, orientation patterns are not consistent between species oracross years. So, is orientation random?

The system is highlyVARIABLE butnot random!

Macroinvertebrate ResultsNumber of individuals from the collector-gatherer feeding guild were lower in wetlands in the reference treatment but other differences minor and not related strongly to buffer treatment.

Results from non-metric multidimensional scaling analyses (not shown) indicted that the major driver of assemblage composition was wetland hydroperiod,not buffertreatment.

Conclusions Vernal pools provide critical habitat for several amphibian species

and a plethora of other species utilize vernal pools for water, food, and as stepping stones to reach other habitats. Organisms that depend on vernal pools are also highly dependent on suitable upland habitat: wetland and upland management should be considered together.

Amphibian population are dynamic and long term data are required to determine effects of management approaches.

During the first few years of this study there is some indication that amphibians breeding in wetlands with 30-m buffers are in worse body condition. The long-term effects of this could be lower productivity.

The 100-m wide clearcuts appear permeable to adult movement but wood frogs do not appear to use clearcuts as summer/fall upland habitat.

Spotted salamander will use underground burrows in clearcuts and their movement behavior is strongly influenced by weather.

Conclusions Orientation direction of amphibians into and out of pools was highly

variable but was consistently not random. Lack of consistent orientation direction among years and between species suggests that in this landscape the a corridor-protection approach would not be suitable.

There is no evidence that the treatments we examined will result in local population loss, but whether this indicates that a 30-m wide buffer is sufficient is still unclear. However, given the data on body condition it seems unlikely that smaller buffers would be adequate.

100-m wide clearcuts appear permeable to movement and the response of amphibians to buffer treatment are partly due to the width of the cuts. Research on response to wider cuts would be necessary to fully understand the connection between upland habitat and long term population viability.

Overall, our results suggest that productive forest management and protection of vernal-pool dependent amphibians can be accomplished but more empirical study on effects of buffer management are needed to assess long term population responses.

Examine longer term and broader scale effects by comparing genetic similarity between the experimental vernal pools and other, near-by pools. By looking at genetic structure, we can start to understand how effectively amphibians are dispersing across the landscape.

To this end, we will collect genetic data from egg masses for all pools adjacent to (within kilometers!) our experimental pools. We will compare the genetic data with known landscape features (e.g., roads, steep hills, lakes) to assess what features promote or impede amphibian dispersal.

?

Continue collecting population and movement measures at experimental wetlands to characterize demographic variability and treatment effects and to inform model development.

Use the results from our long term demographic data and our population genetic data we collect to construct a spatially-explicit decision support tool. Use ground-truthed geospatial data to construct a simulation model that will allow land managers to examine trade-offs between various timber harvesting regimes and amphibian population viability and structure. This will hopefully increase our ability to managed forests for ecologic and economic sustainability.

Veysey, J.S. and K.J. Babbitt. Effects of forest clearcutting on spotted salamander migration and dispersal. American Society of Ichthyologists and Herpetologists Annual Meeting, New Orleans, LA. July 12-17, 2006.

Veysey, J.S. and K.J. Babbitt. Effects of forest clearcutting on spotted salamander migration and dispersal. Eastern CANUSA Forest Science Conference. Québec, Canada. October 19-21, 2006.

Carcagno, E. and K.J. Babbitt. Orientation of vernal pool amphibians in an industrial forest landscape. American Society of Ichthyologists and Herpetologists Annual Meeting. Montreal, Canada. July 23-28, 2008.

Veysey, J., J. Theriault and K.J. Babbitt. Interactive effects of clearcutting and buffer width on vernal pool amphibian breeding demography. American Society of Ichthyologists and Herpetologists Annual Meeting. Montreal, Canada. July 23-28, 2008.

Veysey, J., J. Theriault and K.J. Babbitt. Interactive effects of clearcutting and buffer width on vernal pool amphibian breeding demography. New England Partners in Amphibian and Reptile Conservation Annual Meeting. Rector, PA. August 13-15, 2008.

Veysey, J. and K.J. Babbitt. Vernal pool research at UNH: A precautionary tale. Workshop on vernal pool ecology and documentation. Sponsored by the New England Association of Natural Resources Sciences, New England College,NH Fish and Game Dept., and UNH Cooperative Extension. Henniker, NH. May 29, 2009

Veysey, J. K. Babbitt, J. Purrenhage. An experimental test of buffer width: Impacts on amphibian demography. American Society of Ichthyologists and Herpetologists Annual Meeting. Portland, OR July 22-27, 2009.

Nicole Freidenfelds - Using wood frog (Rana sylvatica) home range and dispersal patterns to predict terrestrial buffer zone requirements in a Maine forest. (M.S. December 2006)

Jessica Veysey - Effects of forest clear cutting on salamander population dynamics and migration. (M.S. December 2006)

Emma Carcagno - Amphibian orientation to vernal pools. (M.S. May 2009)

Joanne Theriault - Effects of forested buffers and wetland characteristics on vernal pool macroinvertebrate assemblages. (M.S. May 2009)

Veysey, J., K.J. Babbitt, and A.C. Cooper. in press. An experimental assessment of buffer width: implications for salamander migratory behavior. Biological Conservation.

Freidenfelds, N.A. and K.J. Babbitt. in manuscript. Post-breeding movement of wood frogs: response to forest buffer size. Target Journal: J. Herpetology, Fall 2009

Veysey, J. and K.J. Babbitt. in prep. Effects of forest buffers on amphibian demographics. Target Journal: Ecology, Winter 2009/2010

Heather Moulton - The effects of upland habitat type on the growth and survival of American toad (Bufo americanus) metamorphs (2006)

Nicole DiManno - Effects of forest buffer width on amphibian use of vernal pools (2006)

Chelsey Faller - Effects of clearcuts on wood frog dispersal rates (2007)

Ashley Green - Fluorescent tracking to observe microhabitat preferences of wood frog (Rana sylvatica) metamorphs in the presence of a clearcut (2007)

Note: These students were supported by UNH Undergraduate Research Fellowships and were required to write proposals, final reports, and present their findings at the annual Undergraduate Research Conference. NSRC and USDA support of the overall project facilitated these student opportunities.