United States
Environmental Protection
Agency
Office of Research and
Development
Washington DC 20460
EPA/600/R-99/066
June 1999
v>EPA PRIMENet
Ultraviolet Radiation/
Amphibian Populations
Research Planning Workshop
February 1-3, 1999
-------
-------
EPA/600;R-99;066
June 1999
PRIMENet
Ultraviolet Radiation/Amphibian
Populations Research Planning Workshop
February 1-3V1999
by:
Peter C. Trenham
Steve A. Diamond
Naomi E. Detenbeck
Gary T. Ankley
U.S. Environmental Protection Agency
Office of Research and Development
National Health and Environmental Effects Research Laboratory
Mid-Continent Ecology Division
6201 Congdon Boulevard
Duluth, Minnesota 55804
Printed on Recycled Paper
-------
Contents
i
Executive Summary • 3
I. Research Objectives 4
A. Agency objectives/roles 4
B. Research hypotheses 5
C. Objectives of research performed under this IAG 6
II. Background °
A. Amphibian monitoring/surveys 6
1. General background '. 6
2. Metapopulation structure and surveys 7
B. Ultraviolet radiation: effects and dosimetry 9
1. Effects of UV on amphibians 9
2. UV dosimetry 10
III. Experimental Design 12
A. Park selection 12
1. Criteria 12
2. Choices 12
3. Research sites in parks of interest 13
B. Species of interest 13
1. Criteria 13
2. Target species 15
C. Survey strategy 15
1. Spatial design • 15
2. Amphibian survey methods 16
a. Retrospective analyses 16
b. Additional data collection 17
3. UV dosimetry measurements 18
D. Manipulative studies 19
]
References 21
Appendices 24
APPENDIX la 24
APPENDIX Ib 25
APPENDIX2 27
APPENDIX 3 2&
APPENDIX 4 29
APPENDIX 5 3i
n
-------
PRIMENet UV / Declining Amphibian Research Planning Workshop
February 1-3, 1999
USEPA Mid-Continent Ecology Division Lab, Dumm, MN
Draft Final Report
Executive Summary
The PRIMENet (Parks Research and Intensive Monitoring of Ecosystems Network) is a
system of 14 national parks, established as index sites for long-term monitoring of environmental
quality and use as outdoor laboratories. From February 1-3* 1999, biologists from various Federal
agencies and academia gathered in Duluth, MN to discuss research to be conducted at PRIMENet
sites under a new interagency agreement (IAG) between the U.S. Environmental Protection Agency
and the U.S. National Park Service. Pursuant to the recommendations of the IAG, participants were
asked to formulate a research program that would address issues of amphibian malformations and
declines and the potential role of ultraviolet radiation. Prior to this meeting biologists from EPA's
Mid-Continent Ecology Division compiled data on amphibian populations and park characteristics,
and determined that Acadia, Smoky Mountains, Rocky Mountains, Glacier, Sequoia-Kings Canyon,
and Olympic National Parks were the feasible sites for this research.
Plenary session presentations and follow-up discussions elaborated the key issues to be
addressed under this IAG, and representatives from each of the 6 candidate sites further described
the relevant characteristics of their parks. Hypotheses to be addressed by this research fell under two
broad headings: 1) the importance of metapopulation dynamics for amphibian surveys and
monitoring; and 2) the potential effects of ultraviolet radiation on amphibians. During the past
decade concerns about amphibian population declines and malformations have increased because
both appear to be on the rise. Metapopulation biology deals with organisms that occupy patchy
habitats where individual patch populations are vulnerable to extinction. Biologists have suggested
that the metapopulation structure of wetland-breeding amphibians may make them susceptible to
declines. In addition, because current survey and monitoring methods assume that individual
wetlands support independent populations, revisions to these protocols may be warranted.
Ultraviolet radiation has also been suggested as possibly contributing to both amphibian declines and
malformations, but many of the relevant experiments have produced ambiguous and apparently
contradictory results. The absence of good UV dose measurements in these experiments, however,
makes comparison virtually impossible.
Although various controlled experiments were discussed, the participants agreed that the
resources of this IAG should be focused on field surveys to close important empirical data gaps.
Because intensive amphibian surveys could not be funded in all parks, a hierarchical design was
suggested that would allow for UV dosimetry work in all 6 candidate parks and amphibian surveys
in a subset of these. Amphibian surveys will be designed to encompass networks of potentially
interacting wetland habitats. Estimates of aquatic UV dose and other habitat associated
measurements will be made for each surveyed wetland. Using this approach, a metapopulation
-------
survey framework will be tested as a means to evaluate the relationship between UV dose and
amphibian habitat utilization in nature; this relationship has yet to be evaluated empirically. Through
the collection of UV and associated habitat data in wetlands at all 6 parks, the potential for
amphibian exposure to damaging levels of solar radiation will also be established. This work will
also contribute to basic knowledge of amphibian distributions within national parks, amphibian
metapopulation structure, UV levels in aquatic environments, and UV measurement technology.
-------
I. Research Objectives
A. Agency objectives/roles
The EPA and NPS's objective in organizing a workshop on solar ultraviolet (UV)
radiation-related amphibian research was three-fold: 1) to review the state of science in UV
dosimetry, amphibian population monitoring, and amphibian UV-effects research, 2) to engage the
Federal and academic scientific communities in planning a strategy for UV-dosimetry and amphibian
effects research at PRIMENet (Park Research and Intensive Monitoring of Ecosystems Network)
sites to be carried out through an existing interagency agreement (LAG) between the Environmental
Protection Agency and National Park Service, and 3) to identify opportunities for collaboration
among Federal agencies engaged in this research now and in the future. The EPA has collaborated
with the NPS to establish a network of sites for intensive long-term monitoring of atmospheric
pollutants and stress-response research on environmental effects related to atmospheric pollutants
(including UV effects) and climate change. The EPA has supported the establishment and
maintenance of Brewer spectrophotometers at 14 PRIMENet sites to measure UV-A, UV-B and
ozone. The NPS contributes to air and deposition monitoring at theses sites with NADP (wet
deposition) and CASTNET (dry deposition) monitoring stations. In addition, meteorological and
visibility (selected sites) parameters are also monitored at the PRIMENet sites.
Effects research currently underway at PRIMENet sites through a past IAG betweenNPS and
EPA includes research on the extrapolation of environmental measurements across natural
landscapes, development of indicators of environmental change, contaminant screening, and
development of amphibian monitoring methods. The latter was rated among the top 10 research
needs by the NPS. Amphibian monitoring work at PRIMENet sites is nested within the NPS
nationwide Inventory and Monitoring (I & M) Program, which involves establishing inventories of
all major natural resources in 250 major resource areas over a 10 year period. The FY2000 DOI
(Department of the Interior) budget initiative includes a substantial addition to the I & M Program
to augment the inventory.
EPA has been involved in UV-dosimetry and UV-effects work with amphibians over the past
few years. In addition, EPA has a strong interest in developing methods of extrapolating dose and
effects measurements from sites to regions, across regions, across species, and from the organism
to population and metapopulation levels, concordant with the goals of EMAP (Environmental
Monitoring and Assessment Program) and CENR (Committee on the Environment and Natural
Resources).
In response to nationwide concern over amphibian declines and malformations, the Secretary
of the Interior has recommended that the DOI take the lead on this issue. A draft budget proposal
for FY2000 has been submitted in support of a 20-year amphibian initiative that includes both
monitoring and research elements. A number of BRD (Biological Resources Division of the U.S.
Geological Survey) researchers who have been active in amphibian research and planning the future
-------
DOI strategy were invited to this workshop to facilitate collaboration among the three Federal
agencies involved and coordination and continuity of research efforts at PRIMENet sites. A full list
of workshop attendees and scheduled presentations and discussion groups are included as
Appendices la and Ib.
B. Research hypotheses
Prior to the February 1999 meeting, representatives from EPA's Mid-Continent Ecology
Division (MED) compiled short lists of hypotheses that this research effort might address (Appendix
2). These hypotheses focused on issues of amphibian exposure to UV radiation and amphibian
population monitoring. During the meeting several of these hypotheses were determined to be
untestable during the 2 to 3 year time-frame of the IAG. Final lists of hypotheses to be addressed
were agreed upon. There was discussion about whether the primary goals of this IAG are research
or monitoring. Due to the short duration of the present IAG and uncertain availability of future
resources, everyone agreed that this effort must focus on answering specific research questions.
However, because monitoring is a concern of both the EPA and NFS, components of this work will
be designed to evaluate the feasibility of novel approaches for future monitoring.
The subgroup discussing population survey issues agreed that wetland-breeding amphibians
theoretically resemble metapopulations, and that survey and monitoring efforts that focus on
networks of potentially interacting subpopulations should be investigated. It was also agreed that,
unless extensive historical amphibian survey data are available, hypotheses relating to extinction and
colonization could not be addressed. Thus, these studies will generally focus on patterns of habitat
occupancy. Four testable hypotheses were identified for further discussion in the research planning
phase:
(1) the distribution of occupied and unoccupied wetland breeding habitats conforms to the
expectations of metapopulation theory;
(2) when effects of wetland isolation and size are accounted for, the probability of wetland
occupancy is inversely related to estimates of UV dose;
(3) rates of local extinction are significantly higher in more isolated and smaller wetlands (only if
historic data exists); and
(4) the spatial distribution of a species is a suitable surrogate for diagnosing temporal population
trends in amphibians.
The group discussing amphibian UV-effects research decided that the hypothesis on which
we should focus this research is: developmental stages exhibit greater frequencies of morphological
abnormalities and/or lower survivorship hi wetlands where higher aquatic doses of UV are received.
There were also suggestions to add immuno-suppression to the list of potential effects, and to
consider UV as only one of many potential stressors in these systems. The group identified UV
exposure issues as a key to evaluating the hypothesis, and listed several related tasks that could be
addressed under the IAG:
(1) measure UV in amphibians habitats;
-------
(2) calibrate habitat-specific measurements to the local Brewer spectrophotometer measurements;
(3) consider a broader light spectrum than just UV-B because other wavelengths may exacerbate or
ameliorate effects; and
(4) characterize amphibian behavior as it relates to potential exposure during sensitive
developmental windows.
C. Objectives of research performed under this IAG
Each of the PRIMENet sites is currently equipped with UV monitoring equipment to gather
systematic data on the potential effect of ozone depletion on ecological resources. In association
with the collection of this data, one of the objectives of the IAG is to investigate relationships
between UV exposure and effects on amphibian populations. The workshop participants were asked
to develop a collaborative research plan that advances techniques for measuring and extrapolating
U V exposures, monitoring and modeling amphibian populations, and assessing associations between
population responses, UV exposure and co-correlated stressors.
-------
II. Background
A. Amphibian monitoring/surveys
1. General background
Inventory and monitoring efforts are costly and require long-term or permanent dedication
of resources. Such efforts should not be initiated without first establishing the ultimate project goals
and the commitment required to achieve these goals. The determination of the number of sites to
be sampled and their locations should be shaped by the goals of the project and not individual biases
or resources. Without the resources to achieve the stated goal(s), the project is virtually certain to
fail.
Monitoring programs are generally able to detect deterministic trends over time. Because
populations fluctuate annually due to weather and random events, monitoring efforts should
concentrate on trends occurring over longer intervals. Ideally the length of the interval over which
trends are reported should be based on empirical observations of natural variability in the species of
interest or a closely related species. In addition the magnitude of the trend considered biologically
significant (e.g., X% per year) should be determined. The reporting interval and magnitude of
interest, along with the chosen alpha and beta levels, are then used to determine the number of sites
required to avoid detecting false trends and failing to detect meaningful trends. Finally the spatial
extent over which trends are of interest (e.g., individual wetland, park-wide) will influence the area
over which survey sites are distributed.
The next consideration is the sampling methodology to be used. It is important to
demonstrate that the methods to be used provide accurate and repeatable estimates of the numbers
of individuals present at a site. Repeatability can be determined by repeated sampling at a single site,
while accuracy is more difficult because it requires that the actual number of individuals present is
known. Because the work we are proposing will be based primarily on presence/absence surveys,
repeatability will be more important than accuracy. Resampling a subset of sites throughout the
season would be useful to assess seasonal variability in detection. Species that cannot be sampled
reliably should be avoided.
2. Metapopulation structure and surveys
As noted above, traditional monitoring programs require long-term commitments to detect
trends hi regional distribution or abundance. In addition, natural features of the life history and
regional distribution of amphibians further complicate these attempts to detect significant shifts in
abundance. However, by altering the amphibian inventory and monitoring paradigm to better fit the
natural history of these animals, we may be able to increase our power to rapidly detect regional
trends and distributional patterns of biological significance.
-------
Many theoretical and empirical studies of organisms that occupy patchily distributed habitats,
such as wetlands, suggest that regular localized extinctions and recolonizations may be a natural
feature of these 'metapopulations' (see Hanski 1998 for additional background). As such, the spatial
configuration of habitat patches may determine the likelihood of regional extinction in networks of
patches, and certainly influences the extinction and colonization probabilities of individual habitat
patches. Pond breeding amphibians utilize inherently patchy aquatic breeding habitats, and studies
have found various species to exhibit characteristics consistent with metapopulation models. Thus,
any attempt to monitor amphibian populations for regional declines must recognize that local
extinctions are probably a natural feature of even healthy amphibian metapopulations.
Traditional estimates of local amphibian abundan.ce and current regional monitoring efforts
have been shown to possess little power to detect meaningful population trends in a timely manner
(Reed and Blaustein 1997; Mossman et al. 1998). Because declines at the level of the
metapopulation would be manifested by local extinctions in excess of colonizations and may produce
detectable spatial patterns of habitat occupancy, surveys that encompass all potential breeding sites
within defined regions may provide a more tractable means for detecting on-going declines. While
extinction and colonization will not be balanced over short intervals, a strength of this type of data
is its potential to reveal landscape-level patterns beginning with a single year's data collection. For
example, if occupancy is primarily driven by random extinction and colonization, occupancy will
be strongly related to pond isolation. However, the influence of other stressors, such as elevated UV
dose, on pond occupancy could also be evaluated using this approach and multiple logistic regression
or other analyses. This approach could be used to screen for the effects of candidate stressors and
to guide future experimental work.
In controlled experiments, researchers have demonstrated that a variety of stressors (e.g., UV-
B, anthropogenic organic compounds) could potentially produce amphibian population declines, but
in most cases the relative distribution of amphibian populations and these stressors has yet to be
established. One exception is the established allotropic distribution of introduced trout and Rana
muscosa in Sequoia-Kings Canyon National Park (Bradford 1989, Bradford et al. 1998). Two
notable amphibian studies have also made explicit use of spatial models to better understand changes
in regional species distributions. First, using an extensive distributional time-series for Rana
lessonae. Sj ogren-Gulve and Ray (1996) used a stepwise logistic regression approach and simulation
modeling to evaluate factors hypothesized to drive observed patterns of population turnover in
ponds. Second, Skelly and Meir (1997) compared observed patterns of amphibian population
turnover in breeding ponds at a Michigan reserve with the patterns predicted if: 1) metapopulation
forces are driving turnover; 2) habitat succession is driving turnover; and 3) turnover occurs at
random. Because some rate of local extinction must be expected in a program designed to monitor
species exhibiting metapopulation characteristics, and especially one designed to detect declines,
protocols should be designed such that patterns of habitat occupancy or turnover can be used to test
alternative models such as described above.
Current amphibian monitoring protocols focus on roadside calling surveys, where individual
-------
wetlands separated by a minimum of 0.5 miles are identified and visited several times throughout
the spring and summer. These surveys are well suited to determining broad species distributions and
detecting changes occurring over large regions and long periods of time. However, data collected
using this approach are poorly suited to evaluate factors correlated with regional shifts because
potentially crucial local features of individual wetlands or wetland networks are largely ignored. If
an objective of surveys and monitoring is to understand the factors influencing regional patterns of
habitat occupancy and turnover, surveys should include additional data collection, such as isolation
distance, water chemistry, and vegetation characteristics. While this approach is novel and
unproven, the above examples suggest that data collected using a geographically comprehensive
survey framework will allow improved power to evaluate factors suspected to drive current
distributions. The power of geographically comprehensive surveys for improved detection of trends
over time will have to be evaluated using simulated data, because of the absence of sufficient data
sets and the immediate practical interest in the power of this approach. With the proposed increase
in DOI resources for amphibian monitoring and research, a new sampling approach 'that facilitates
regional inventories and pattern analysis would be a valuable tool.
The implementation of geographically comprehensive survey and monitoring protocols
would first require the identification of survey regions in which all patches would be sampled. Patch
sampling would involve determining amphibian presence or absence and collecting data on
associated habitat characteristics and potential stressors. Regions could be defined by watershed or
other boundaries likely to limit most individuals to within-region movements. Densities of
potentially suitable amphibian breeding habitat could be identified from wetland inventory maps for
selection of regions optimal for surveys. Because habitats are only surveyed for presence or absence,
and the effects of multiple factors (e.g., isolation and UV dose) will be analyzed simultaneously, a
large number of patches (e.g., >100) will need to be sampled to attain sufficient statistical power.
B. Ultraviolet radiation: effects and dosimetry
1. Effects of UV on amphibians
Solar ultraviolet (UV) radiation, and more specifically the UV-B region of the spectrum (see
below for further definition), has been cited as a potential contributor to the recent apparent declines
of, as well as malformations in, amphibian taxa. There are several factors associated with UV
radiation that make such hypotheses particularly worthy of investigation. First, there is little
argument that the expected chlorofluorocarbon-caused reduction in the stratospheric ozone layer has
occurred as predicted and has resulted in the expected significant increases in terrestrial UV-B
intensity (UNEP, 1998). Second, the capacity of UV-B radiation to directly impact biological
macromolecules, including DNA, in a variety of taxa has been well documented. Third, early life
stages of all taxa, and especially species in which significant metamorphosis occurs, are particularly
susceptible to the damaging effect of UV-B radiation. This susceptibility is associated with the
potential for greater exposure to sunlight, the potential for UV penetration through the entire embryo
or early larvae, the rapid replication and differential expression of DNA, and the accelerated
10
-------
production of potential UV- receptor biomacromolecules during these life stages. And fourth, other
global and regional impacts not directly associated with atmospheric transmission of UV can
significantly impact the UV dose received in aquatic environments (Schindler et al. 1996). Many
of these impacts are associated with alteration in the flux of DOC, by far the most effective UV-
attenuating component of natural waters (Williamson et al. 1996). Other impacts include subtle
alterations in nutrient loads (and concomitant algal growth), loss or alteration of riparian vegetation,
sedimentation, etc.
There have been a number of laboratory and field studies examining the effects of UV
radiation on survival and development of amphibians. There is no question that it is possible to elicit
reduced survival and a wide range of malformations in early embryonic stages exposed to UV (e.g.,
Worrest and Kimeldorf 1975;1976; Blaustein et al. 1994; 1995; 1997; Grant and Licht 1995; Long
et al. 1995; Hays et al. 1996). Recent studies have shown that UV light can also elicit adverse
developmental (specifically limb malformations) effects in later stages of metamorphosis in
amphibians (Ankley et al. 1998; in review). The major question, therefore, is not whether UV light
can adversely affect frogs, it is whether or not current UV intensities, or projected changes associated
with global climate alterations, adversely affect amphibians. Currently at the heart of this
controversy are the results of a series of manipulative studies, conducted predominantly in the field,
which have yielded ambiguous or conflicting data. Specifically, Blaustein and coworkers (e.g.,
Blaustein et al. 1994; 1995; 1997) have published several papers suggesting that current levels of
UV radiation (specifically UV-B) are sufficient to reduce survival of at least some amphibian
species. By extension, these UV effects would only be expected to worsen, at least in the short-term
(UNEP 1998). Other investigators, however, have not been able to replicate the results of Blaustein
et al. (e.g., Grant and Licht 1995; Ovaska et al. 1997; Corn 1998). Although some of the seeming
among-study contradictions may be explainable solely based on experimental design (e.g., timing
and duration of exposure), the largest uncertainty in terms of comparison is due to dosimetry issues.
Basically, the majority of studies conducted to date with UV and amphibians, in particular those in
the field, have done so little in terms of dosimetry, either from a quantitative and qualitative
perspective, that it is virtually impossible to compare them to one another, or to project the
implications of their results to other settings (including future scenarios of potentially increasing UV
intensities). This general shortcoming is further exacerbated by a lack of reasonable UV dosimetry
data for natural amphibian habitat.
The research outlined in this report will directly address the issue of effects extrapolation in
the context of UV dosimetry in two ways. First, planned laboratory and natural sunlight exposure
studies, in support of the research outlined herein, will emphasize the appropriate measurement of
dose. This work, although not directly supported by this IAG, should help put the data collected into
a realistic context from the perspective of dose-response relationships; studies will be conducted by
both the EPA (MED) and the BRD (Columbia). Second, work supported by this IAG and other
research sponsored by EPA (in collaboration with the University of Minnesota), should help address
the issue of UV dosimetry relative both to amphibian habitat and larval behavior.
11
-------
2. UVdosimetry
The question of whether increasing reports of amphibian declines or malformations are
associated with UV intensity can only be addressed at the habitat level. Regional monitoring of
ozone layer depth and/or UV-B intensity, while useful in characterizing global UV fluctuation and
perhaps hi providing baseline intensity information, provides little habitat-specific intensity data.
For example, regional monitoring data might suggest little overall increase in UV intensity, yet even
small alterations hi DOC concentration or character (due to local or global effects) could result in
an order-of-magnitude increase in total UV dose in specific water bodies (Schindler et al. 1996). The
potential for significant temporal fluctuation associated with all of the factors discussed above
exacerbates the need for habitat-specific UV characterization and quantification.
The science of dose estimation for solar actinic radiation (SAR), or more specifically UV
(ultraviolet radiation: UV-B = 280 nm to 320 nm, UV-A = 320 nm to 400 nm), is largely
undeveloped for aquatic habitats. In general, attempts to estimate doses of UV or SAR have been
based either on projections of single solar radiation measurements and/or calculations based on
global location, season, tune of day, etc. These estimations fall far short of accurately quantifying
UV dose because several key factors, including weather, attenuation characteristics of the water
column, and animal behavior, are not quantitatively considered.
The estimation of aquatic UV dose, although complex, can be considered to consist of two
major components: 1) terrestrial UV dose; and 2) transmission of light within the aquatic
environment. Terrestrial irradiance (the total intensity of direct and reflected radiation delivered to
a horizontal surface) is influenced by geometric considerations including global location, season and
the time interval considered. In addition more variable factors, including attenuation by stratospheric
ozone, other atmospheric components, weather and local shading, can dramatically influence
irradiance. The effects of location and tune are easily modeled using trigonometric calculations that
predict the angle-of-incidence of sunlight and the associated dispersion of radiation over a horizontal
surface. The results of these initial calculations can be modified by more complex algorithms that
estimate the effects of various atmospheric factors. However, these variable factors are much more
difficult to model with confidence, particularly when very specific local habitats are considered.
Historical data on terrestrial irradiance are limited to a few locations and/or very short time periods.
These data, although of use in calibrating or testing irradiance models, have little direct application
for the estimation of UV dose at PRIMENet sites.
The second component of UV dose estimation, transmission of light in aquatic environments,
adds significant complexity to UV dose estimates for specific taxa. In addition to the factors
affecting terrestrial solar transmission, the aquatic environment introduces several highly variable
factors that can dramatically alter the intensity of irradiance reaching any given location or micro
habitat. These factors include dissolved organic material (DOM or DOC), suspended inorganic and
organic material, and shading vegetation. The wavelengths of radiation that comprise sunlight
spectra are affected differentially by all of these factors. Most important for UV dose estimation
12
-------
in aquatic habitats is DOC, which depending on its constituency (fulvics, humics,: tannins, etc.) can
vary significantly in its ability to attenuate UV radiation. Shading and reflection can increase the
ratio of UV relative to visible light, and the intensity of shorter wavelengths tends to be diminished
more significantly at low sun-angles. For example, at dawn and dusk UV-B is undetectable yet
visible radiation can be sufficient for reading.
Finally, not all wavelengths of light are equally effective at eliciting a given response, either
physiological or behavioral. For example, Setlow's DNA dose, a quantitative evaluation of the
potential for DNA damage, is determined by the absorption of radiation by DNA which occurs
largely in the UV-B range of terrestrial solar radiation (the DNA absorption maxima is at 254 nm
and tapers steeply to essentially zero at 310 nm). Even small changes in the attenuation of UV-B
by DOC or other factors can have order-of-magnitude effects on the calculated dose, depending on
the spectrum of the response being considered. Hence, dose estimation should minimally include
an initial quantification of the light spectra present. Also adding to this complexity is the need to
characterize specific taxa as to their potential for exposure to sunlight. For example, factors
influencing exposure risk will include the amount of time a species or life-stage spends in open water
during high-intensity events (e.g., mid-day, mid-summer, clear-sky conditions), and any outer tissues
or other materials that might shield them from radiation (e.g., shell, carapace, or constructed
dwelling).
13
-------
HI. Experimental Design
A. Park selection
1. Criteria
From descriptions provided by park representatives, workshop attendees compiled
information on the characteristics of the 14 PRIMENet sites to identify those best suited for studies
relevant to: 1 Investigating habitat-specific UV levels; 2) Considering issues of amphibian decline;
3) Conducting manipulative UV effects research, and 4) Testing a geographically-comprehensive
survey protocol. The criterion for UV studies was a range of UV-exposures which might be
provided by the existence of an elevation, DOC, or vegetation gradient across aquatic habitats.
There were two criteria identified for deciding whether amphibian declines should be addressed in
the national parks. One was whether there was evidence that park amphibian populations are
presently less abundant or widespread than they have been historically. The other was whether
species present hi the parks were in apparent decline in other non-park locations. Criteria for effects
research included the presence of species of interest that metamorphose during their first summer
and a gradient of UV exposure within the park. Finally, because we are interested in investigating
the use of techniques that require patch-based data, the presence of species that utilize discrete
wetlands as breeding habitat is critical. Additional criteria for geographically comprehensive surveys
included the presence of several detectable and widespread species of wetland breeding amphibians,
at least 100 potentially suitable discrete wetland breeding habitats, and the existence of GIS
landscape coverages (e.g., elevation, hydrography,'trails, roads, land-cover, historic amphibian
occurrences) for the park. Finally, wherever present, species overlap between parks was noted.
2. Choices
Prior to the February 1999 meeting eight of the PRIMENet sites were eliminated from further
consideration, based on the above criteria. Denali and Hawaii Volcanoes National Parks were
eliminated due to a lack of multiple native species. Everglades National Park was eliminated due
to a lack of sufficient topographical relief such that individual discrete wetlands could not be easily
identified. Shenandoah and Virgin Islands National Parks were eliminated due to a lack of sufficient
numbers of wetlands. Big Bend and Canyonlands National Parks were eliminated due to the regional
dominance of amphibian species that are difficult to survey confidently due to their unpredictable
activity patterns. Theodore Roosevelt National Park was eliminated due to a lack of basic survey
data to indicate how many wetlands and what species are present.
At the meeting, representatives from Acadia, Great Smoky Mountains, Rocky Mountain,
Glacier, Sequoia-Kings Canyon, and Olympic National Parks presented additional background
relevant to the site selection criteria (summarized in Appendix 3). Based on these presentations and
follow-up discussions, Great Smoky Mountains and Sequoia-Kings Canyon National Park were
determined to be less suitable for implementation of geographically comprehensive amphibian
14
-------
surveys. Great Smoky Mountains National Park was eliminated due to a marginal number of
wetland habitats, and Sequoia-Kings Canyon National Park was generally considered to be less
suitable due to suspected aerial loadings of anthropogenic organic chemicals that might confound
effects of interest. However, the severity and extent of contamination is not established, and several
representatives thought other parks may also receive similar inputs.
3. Research sites in parks of interest
Study sites within the parks will be identified using wetland maps and records of past
amphibian observations to identify suitable areas. The primary concern for the survey component
of the project is that at least 100 potentially suitable wetland breeding habitats can be located and
sampled. In some parks this may require visiting virtually all available wetlands in the park, while
elsewhere these surveys will only encompass a fraction of the available wetlands. In cases where
all wetlands cannot be surveyed, parks will be subdivided based on watershed or other boundaries,
and as many of these subregions as possible will be surveyed. Sampled wetlands should be
distributed along gradients suspected to influence aquatic UV-dose (e.g., elevation, DOC,
vegetation), and areas exhibiting an isolation gradient should be favored.
B. Species of interest
1. Criteria
We compiled basic information on all amphibian species known to occur at PRIMENei sites
(Appendix 4), and species criteria varied depending on the research component under consideration.
For geographically comprehensive amphibian surveys, species criteria were that discrete wetlands
are the preferred breeding habitat, that some life stage is readily detectable in this habitat for much
of the summer, and that species have a wide distribution within a park. Frogs and toads were noted
as preferable over most salamanders, as were early breeders and species with shorter life spans
because they will be less likely to skip breeding arid yield spurious annual absences. Table 1 lists
the pond breeding species occurring at each of the five parks still under consideration for survey
work. In addition, species that are documented as declining in parks or other areas will be given
priority when the suite of species to be surveyed is detennined for each park. For effects research,
the primary criteria for assessing potential species of interest was a life cycle conducive to
manipulative experimentation and monitoring of the different life stages (i.e., metamorphoses during
first summer).
15
-------
Table 1. Wetland-breeding amphibians at PRIMENet sites. Asterisks indicate species noted
by participants as particularly feasible for proposed survey efforts, due to relative ease and
confidence of sampling. Footnotes indicate species noted as especially troublesome for
reasons explained in the footnotes.
Species
Spotted salamander*
Red-spotted newt*
Four-toed salamander
Spring peeper*
Gray treefrog*
American toad*
Leopard frog2
Pickerel frog2
Bullfrog
Green frog*
Wood frog*
Tiger salamander*
Chorus frog*
Boreal toad*
Long-toed salamanders
Columbia spotted frog
Pacific treefrog*
Yosemite toad
Mtn. yellow-legged frog*
Northwestern salamander
Red-legged frog3
Cascades frog
Rough-skinned newt*
Acadia
X
X
X
X
X
X
X
X
X
X
X
Rocky
Mountain
X
X
X
X
Glacier
X
X
X
X
Sequoia-
Kings
Canyon
X
X
X
X
X
Olympic
X
X
X
X
X
X
X
X
1 difficult to detect in favored bog habitats
2 hard to locate
3 weak call makes them hard to detect at low densities
2. Target species
Target species will have to be identified on a site by site basis using the criteria discussed
above, and surveys will be optimized for these species. However, information will be gathered on
all amphibian species detected during surveys. Wherever possible data will be collected on species
that overlap between parks, but species overlap was not identified as a critical characteristic for park
selection. Species potentially of interest due to their presence at more than one park are boreal toads
16
-------
(Bufo boreas) and wood frogs (Rana sylvatica). These two species are also of particular interest
because their eggs and tadpoles are relatively easy to detect.
C. Survey strategy
The basic goals of the proposed surveys are: 1) To assess the logistic efficiency and
diagnostic power of geographically comprehensive amphibian surveys; 2) To gain a better
understanding of the doses of UV-B radiation likely to be received by embryonic and larval
amphibians; 3) To determine the factors influencing aquatic UV dose, and; 4) To determine if current
amphibian distributions are consistent with the negative impacts associated with elevated UV
exposure. To achieve these goals we have proposed a hierarchical framework whereby some level
of habitat-specific UV dosimetry will occur at all 6 PRIMENet candidate sites, while more time- and
labor-intensive efforts will occur at some of the sites identified as suitable for geographically
comprehensive amphibian surveys.
The workshop participants discussed time-lines for the work to be conducted under this JAG
and through in-kind cooperation, extending from F Yl 999 through F Y2000 or F Y2001. Opportunities
for collaboration with investigators working on related research in the parks (Appendix 5) we also
identified. It was also concluded that the IAG funds should be used to support amphibian surveys
and habitat-specific UV measurements in the parks, while related experimental studies would be best
addressed through in-kind cooperation between interested parties. Project schedules were not
discussed in detail, but some general targets were suggested. Dr. Corn suggested that retrospective
spatial analyses of recent park amphibian inventories be completed before further field surveys are
initiated. He suggested that the summer of FY1999 be used for park and study site selection within
parks and possibly limited field surveys, while FY2000 be used for more extensive surveys
integrated with UV analyses.
1. Spatial design
The primary reason for collecting geographically comprehensive survey data is that habitat
isolation has been repeatedly demonstrated to influence wetland colonization and extinction, and as
a result impacts occupancy patterns in amphibian systems. Thus, for each wetland aprimary variable
of interest must be the distance to the nearest occupied site(s). Although factors such as water
quality may also influence habitat use, isolation effects must be accounted for or controlled for, less
their effects obscure significant patterns of greater interest. Because standardizing isolation is
difficult or impossible in most natural systems, techniques such as multiple logistic regression are
used to assess the effects of isolation and other factors on habitat occupancy. Thus, it is critical that
amphibian presence/absence and habitat/stressor characteristics be established for all potentially
interacting wetlands within a given region. The scale at which wetlands support potentially
interacting populations is difficult to establish and varies between species. However, based on
published records, wetlands separated by more than 3 krn are probably unlikely to exchange frequent
migrants for most species. Certain landscape features, such as large lakes and mountains, may also
17
-------
represent barriers to movement and constrain likely interactions.
Because national parks will often be too large to attempt surveys of all park wetlands, and
some areas may be prohibitively difficult to access, it was suggested that each park be divided into
multiple potential study regions based on watershed boundaries. Potentially suitable wetland
breeding habitats in each of these watershed regions would be located using National Wetland
Inventory (NWI) maps and aerial photographs, so that watersheds could be categorized based on
number and density of wetlands, elevation, and accessibility. Because the goal of this project is to
answer a research question (i.e., does aquatic UV dose have a negative effect on wetland occupancy),
and not an unbiased amphibian inventory or monitoring program, site selection for this project need
not be completely randomized. Areas known to be completely devoid of species of interest will be
avoided, as will areas supporting too few wetland habitats to produce sufficient data return for effort
invested (i.e., < 10). Watersheds identified as "reasonably accessible" and containing at least 10
wetland patches and species of interest will be targeted for surveys. These watersheds will then be
divided into categories based on elevational ranges, so that sampling can be stratified among
elevation categories equally. The general consensus was that, as long as each region contains at least
one occupied patch and all potentially interacting regional habitat patches are surveyed, data from
non-contiguous regions within a given park could be pooled with little effect. The potential impacts
of data pooling in this way should be considered hi greater detail.
Concern was voiced about making these survey techniques more directly relevant to unbiased
monitoring efforts. One suggestion was to use a sub-sampling technique analogous to the 'point-
quarter' technique utilized hi vegetation surveys. Another possibility would be to randomly choose
individual wetlands and then survey all wetlands within some set distance. A modeling effort would
probably be best suited to evaluating the utility of various site selection and sampling schemes.
2. Amphibian survey methods
a. Retrospective analyses
In recent years, many national parks have completed extensive inventories of the amphibian
populations within their boundaries (Table 2). In many cases these surveys encompassed essentially
all potential wetland breeding sites within entire parks or large regions of the parks. While identical
survey methods were not utilized between all parks, these data sets provide opportunities to begin
to evaluate the effects of wetland isolation and other landscape features on habitat patch occupancy.
We are beginning to analyze the published inventory data from Rocky Mountain National Park
(Corn, et al. 1997). Amphibian inventory data will be included as layers on a GIS.map of the park
to facilitate spatial analysis of amphibian and landscape data. We will send out a letter of inquiry
to all other parties hi possession of similarly extensive amphibian inventory data from PRIMENet
sites, offering to collaborate with researchers on similar analyses. To these collaborations we bring
the MED's permanent GIS staff and expertise with landscape and metapopulation level analyses.
18
-------
Table 2; Recent inventories of PRIMENet site amphibians
Park
Acadia
Great Smoky
Rocky Mountain
Glacier
Sequoia-Kings Canyon
Olympic
Researcher(s)
Ken Dodd, Jim Petranka
Steve Corn
Leo Marnell
Gary Fellers, Roland Knapp, David Bradford, Kathleen Matthews
Bruce Bury and Mike Adams
b. Additional data collection:
To maximize the ability to draw reliable comparisons between parks, future surveys should
utilize standardized methods and levels of effort, both within and between parks. A variety of survey
techniques are utilized to survey for amphibians, but unfortunately no one technique is ideal for
detecting all species. Survey techniques will be focused on those that are most useful in sampling
lentic wetland habitats (e.g., lakes, ponds, marshes, swamps). The primary variable of concern for
each wetland is the presence-absence determination for each of the species occurring in the park.
Although we are interested in any evidence of a species' occurrence and any such data should be
recorded, evidence of local breeding activity is the information of choice. Survey techniques
suggested by panel members were visual surveys, calling surveys, sweep net sampling, and traps.
Calling surveys were considered useful but insufficient, due to the often weak link between these
observations and the presence of breeding populations. Traps were also considered potentially useful
but insufficient because they passively sample only small portions of wetlands. Larval surveys
utilizing a combination of visual searches and sweep net samples were generally considered the most
reliable method for establishing the presence of breeding populations. Repeated sampling at a subset
of sites was suggested as a means of establishing the probability of false negatives in larval surveys,
and how detectability varies between species.
At some point during visits to each pond the team will also collect water samples, perform
other necessary UV measurement duties (see below), and measure a variety of habitat characteristics.
Because amphibians utilize aquatic and terrestrial habitats, and both may be limiting, aspects of each
of these environments will need to be characterized. A suite of water quality variables will be
measured including temperature, turbidity, hardness, etc. Water temperature may be among the most
important habitat variables measured because it may be related to UV dose and can dramatically
influence larval growth and development. If contaminants are considered an issue to be addressed
at specific park(s), the team may also need to collect samples for this purpose. For each pond the
exact location and dimensions should be accurately determined using a global positioning system
(GPS).
3. UV dosimetry measurements
19
-------
The overall goal of the initial two-year monitoring effort should be to characterize the range
and variability ofsolar radiation flux among all wetlands selected for geographically comprehensive
amphibian surveys. These data will then be used to test for correlation between estimated solar
radiation dose and the presence or absence, and possibly the relative density, of amphibians among
patches. In addition, the range ofsolar radiation flux among measurement locations will provide
insight into the predictive usefulness of the Brewer spectrophotometers located at PRIMENet sites.
Two major points should be emphasized relative to this effort. First, solar radiation
measurement at the selected PRIMENet sites should be guided by the amphibian survey efforts. If
possible, all wetlands surveyed for amphibians should also be characterized relative to their solar
radiation exposure, and then, time allowing, additional sites should be examined. Second,
characterization of additional factors including animal behavior, seasonal variability in UV
attenuation and shading factors would be required to confidently assess UV dose received by
developing amphibians. Therefore, due to limited funding and time, these data will provide only
preliminary estimates of UV dose.
Solar radiation measurements and habitat characterization will involve several procedures
to be detailed in later technical documents. In general, water samples from each patch will be
collected and returned to the EPA's MED Laboratory in Duluth, Minnesota for spectrophotometric
quantification of spectral attenuation capacity. Field-filtered (0.45 um) water samples need to be
shipped to MED within 5 days hours of collection. Mid-day, full-sun extinction data will be collected
using a UV-A/UV-B broad waveband radiometer. Both the water samples and the extinction data
will be collected from a central, unshaded (if possible) location within each wetland. A photographic
or sketched record of wetland coverage by aquatic vegetation and algae will be generated for each
patch. Gross shading (large vegetation, physical geography, etc.) will be characterized by estimating
the angle of inclination at several compass points from a central location within each patch. Given
sufficient time, additional efforts could include selected-patch UV-B intensity mapping and day-long
extinction quantification using polysulfone dosimeters, and patch spectral attenuation and dosimetry
using a portable' spectrometer (this would necessitate site visits by an EPA MED scientist).
D. Manipulative studies
Given the resources needed for baseline UV dosimetry work and amphibian surveys at select
PRIMENet sites, it is not envisioned that there will be support for manipulative field studies via this
IAG. However, this type of experimentation could provide a critical link between field-collected UV
dosimetry and presence/absence data in terms of developing causal stressor-response associations.
Therefore, if further resources become available, these types of studies should be considered. The
types of manipulative studies that seem reasonable fall into two broad classes: 1) Those in which a
given habitat containing amphibians is systematically altered so as to change the UV dose, and; 2)
Those in which subsets of animals from a given "cohort" are transplanted to multiple locations
receiving differential UV exposures. Of course, studies that have attributes of both types of
experiments could also be conducted. In the first case, UV dose to systems could be varied through
20
-------
"natural" alterations, such as increasing or decreasing shading, or modified through the use of
wavelength-specific or neutral density filters. The whole system (e.g., wetland) could be modified,
or replicate experimental units within the system could be subject to UV modifications. The latter
of these alternatives has been commonly utilized (e.g., Blaustein et aL 1994; 1995; 1997; Corn
1998). Transplantation studies, although potentially more complicated, could also be very useful in
defining cause-effect relationships. Transplantations would be particularly attractive in parks where
a natural gradient of UV intensity can be identified that is consistent with the distribution of the
amphibian species under consideration. In any type of manipulative study, an important aspect of
data collection would include an evaluation of animal behavior in the context of UV exposure.
Further, these studies would need to be conducted with careful attention given to continual on-site
solar exposure estimation.
Concurrent with the field research conducted at the National Parks there will be controlled
experimentation concerning the effects of UV on amphibian survival and development at both EPA
(MED) and BRD (Columbia) facilities. Although not directly supported by the EPA/NPS LAG, this
work will be conducted so as to complement, to the fullest extent possible, studies at the park sites.
For example, building upon the observations of Ankley et al. (1998; in review), EPA scientists will
continue to investigate the effects of UV on survival and development of native North American
Ranids. One emphasis of these studies will be on hindlimb dysmorphogenesis, but an important
component of the work also will be evaluation of the comparative early life-stage sensitivity of
several anuran species to UV light. For example, comparative studies have been/will be conducted
with the northern leopard frog, green frog, mink frog and wood frog. Additional species may be
added depending upon site and species selection associated with the PRIMENet studies. The point
of these studies is to develop robust dose-response relationships, using both simulated (laboratory)
and natural sunlight, for survival through metamorphosis and specific developmental endpoints.
Dose-response data will be collected with careful attention to UV exposure characterization, such
that these dose-response relationships can be directly extrapolated to field collected UV exposure
data.
The natural sunlight experimentation will be conducted using the exposure facility described
by Ankley et al. (in review). This system enables a large degree of flexibility with respect to sunlight
dose (both from a quantitative and qualitative perspective), while maintaining a highly controlled
environment in all other respects (e.g., temperature, general water quality, influence of external
variables such as precipitation). Further, the system design facilitates performance of precisely
replicated studies in terms of homogenous sunlight dose across a potentially large number (>40) of
experimental units. Finally, real-time, continuous UV monitoring provides unique capabilities in
terms of dose estimation in simulated field settings. Given these characteristics, the system provides
unique opportunities, not only for dose-effects research, but for behavioral studies focused upon
defining potential UV dose to amphibians as a function of key, potentially-sensitive stages of
development.
21
-------
References
Ankley, G. T., J. E. Tietge, D. L. DeFoe, K. M. Jensen, G. W. Holcombe, E. J. Durban, and S.A.
Diamond. Effects of methoprene and ultraviolet light on survival and development ofRana
pipiens. Environmental Toxicology Chemistry 17:2530-2542 (1998).
Ankley, G. T., J. E. Tietge, G. W. Holcombe, D. L. DeFoe, S. A. Diamond, K. M. Jensen, and S.
J. Degitz. Induction of hindlimb malformations in Ranapipiens by artificial ultraviolet light
and natural sunlight, (hi review.)
Blaustein, A. R., P. D. Hoffman, D. G. Hokit, J. M. Kiesecker, S. C. Wells, and J. B. Hays. UV
repair and resistance to solar UVB in amphibian eggs: A link to population declines?
Proceedings of the National Academy of Sciences of the United States of America 91:1791-
1795 (1994).
Blaustein, A. R., B. Edmund, J. M. Kiesecker, J. J. Beatty, and D. G. Hokit. Ambient ultraviolet
radiation causes mortality hi salamander eggs. Ecological Applications 5:740-743 (1995).
Blaustein, A. R., J. M. Kiesecker, D. P. Olivers, and R. G. Anthony. Ambient UV-B radiation
causes deformities in amphibian embryos. Proceedings of the National Academy of Sciences
of the United States of America 94:13735-13737 (1997).
Bradford, D.F. Allotropic distribution of native frogs and introduced fishes in high Sierra Nevada
lakes of California: Implications of the negative effect of fish introductions. Copeia
1989:775-778 (1989).
Bradford, D. F., S. D. Cooper, T. M. Jenkins, K. Kratz, O. Samelle, and A. D. Brown. Influences
of natural acidity and introduced fish on faunal assemblages in California alpine lakes.
Canadian Journal of Fisheries and Aquatic Sciences 55:2478-2491 (1998).
Corn, P.S. Effects of ultraviolet radiation on boreal toads in Colorado. Ecological Application
8:18-26(1998).
Corn, P. S., M. L. Jennings, and E. Muths. Survey and assessment of amphibian populations in
Rocky Mountain National Park. Northwestern Naturalist 78:34-55 (1997).
Grant, K. P., and L. E. Licht. Effects of ultraviolet radiation on life-history stages of anurans from
Ontario, Canada. Canadian Journal of Zoology 73:2292-2301 (1995).
Hanski, I. Metapopulation dynamics. Nature 396:41-49 (1998).
Hays, J. B., A. R. Blaustein, J. M. Kiesecker, P. D. Hoffman, I. Pandelova, D. Coyle, and T.
22
-------
Richardson. Developmental responses of amphibians to solar and UVB sources:
comparative study. Photochemistry and Photobiology 64:449-456 (1996).
A
Long, L. E., L. S. Saylor, and M. E. Soule. A pH/UV--B syhergism in amphibians. Conservation
Biology 9:1301-1303 (1995).
Mossman, M. J., L. M. Hartman, R. Hay, J. R. Sauer, and B. J. Dhuey. Monitoring long-term trends
in Wisconsin frog and toad populations. In: M. J. Lannoo, ed., Status and Conservation of
Mid-Western Amphibians. University of Iowa Press, Iowa City, LA (1998).
Ovaska K., T. M. Davis, and I. Novales Flamarique. Hatching success and larval survival of the
frogs Hyla regilla and Rana aurora under ambient and artificially enhanced UV radiation.
Canadian Journal of Zoology 75:1081-1088 (1997).
Reed, J. M., and A. R. Blaustein. Biologically significant population declines and statistical power.
Conservation Biology 11:281 -282 (1997).
Schindler, D. W., P. J. Curtis, and M. P. Stainton. Consequences of climate warming and lake
acidification for UV-B penetration in North American boreal lakes. Nature 379:705-708
(1996).
Sjogren-Gulve, P., and C. Ray. Using logistic regression to model metapopulation dynamics: large-
scale forestry extirpates the pool frog. In: D.R. McCullough, ed., Metapopulations and
Wildlife Conservation, Island Press, Washington, DC. 111-137(1996).
Skelly, D. K. and E. Meir. Rule-based models for evaluating mechanisms of distributional change.
Conservation Biology 11:531-538 (1997).
UNEP (United Nations Environment Program). Environmental effects of ozone depletion: 1998
Assessment, Draft Report (1998).
Williamson, C. E., R. S. Stemberger, D. P. Morris, T. M. Frost, and S. G. Paulsen. Ultraviolet
radiation in North American lakes: attenuation estimates from DOC measurements and
implications for planktonic communities. Limnology and Oceanography 41:1024-1034
(1996). .
Worrest, R. C., and D. J. Kimeldorf. Photoreactivatiori of potentially lethal, UV-induced damage
to boreal toad (Bufo boreas boreas) tadpoles. Life Sciences 17:1545-1550 (1975).
Worrest, R. C., and D. J.Kimeldorf. Distortions in amphibian development induced by ultraviolet-B
enhancement (290-315 nm) of a simulated solar spectrum. Photochemistry and
Photobiology 24:377-382 (1976).
23
-------
APPENDIX la
List of attendees
Research Planning for Amphibian Monitoring and Assessment at National Park Service/USEPA Index Sites (February
1-3, 1999)
Dr. Michael J. Adams
USGS Biological Resources Div.
3080 SE Clearwater Drive
Corvallis, OR 97333
541-754-4382
Dr. David Bradford
US EPA NEERL Characterization Research Div.
944 East Harmon
P.O. Box 93478
Las Vegas, NV 89193-3478
702-798-2681
Dr. Stephen Com
USGS Midcontinent Ecological Science Ctr
Aldo Leopold Wilderness Research Inst.
PO Box 8089,790 E. Beckwith Avenue
Missoula, MT 59807
406-542-4191
Ken Czarnowski
Physical Sciences Coordinator
Rocky Mountain National Park
Division of Resources Mgrat and Research
EstesPark, CO 80517
Dr. Sam Droege
USGS-BRD, PWRC
12100 Beech Forest Dr
Laurel, MD 20708-4038
301-497-5840
Patti Happe
Olympic NP
600 East Park Ave.
Port Angeles, WA 98362
Dr. Edward Heithmar
US EPA NEERL Characterization Research Div.
944 East Harmon
P.O. Box 93478
Las Vegas, NV 89193-3478
702-798-2626
Dr. Edward E. Little
USGS, Biological Resources Division
Columbia Environmental Research Center
4200 New Haven Road
Columbia, MO 65201
David Manski
Chief, Division of Resource Management
Acadia National Park
P.O. Box 177
Bar Harbor, ME 04609
Dr. Bernard Shanks (MS-300)
US Dept. of Interior
USGS Biological Resources Division
12201 Sunrise Valley Drive
Reston,VA20192
Dr. David Skelly
School of Forestry and Environmental Studies
Yale University, 370 Prospect Street
New Haven, CT 06511
203-432-3603
Dana Soehn
Great Smoky Mountains Nat'l Park
1314 Cherokee Orchard Road
Gatlinburg, TN 37738
Dr. Kathy Tonnessen
NPS-AIR
P.O. Box 25287
Denver, CO 80225-0287
Dr. Barbara Walton
US-EPA NHEERL MD-87
Research Triangle Park, NC 27711
919-541-7776
Harold W. Werner
Sequoia-Kings Canyon National Park
Three Rivers, CA 93271
559-565-3123
US EPA Mid-Continent Ecology Division (6201 Congdon Blvd., Duluth, MN 55804,218-529-5000)
Dr. Gary Ankley Mr. Gary Holcombe
Dr. Steven Bradbury Dr. Joseph Tietge
Dr. Steven Hedtke Dr. Anett Trebitz
Dr. Naomi Detenbeck Dr. Peter Trenham
Dr. Steve Diamond Ms. Sharon Batterman
24
-------
APPENDIX Ib
Schedule of presentations / activities:
Research Planning for Amphibian Monitoring and Assessment at National Park Service/USEPA Index Sites
February 1-3, 1999
USEPA/ORD Mid-Continent Ecology Division
6201 Congdon Blvd.
Duluth, MN
Large Conference Room
AGENDA
Monday, February 1
1 pm Introductions and overview of agency and workshop objectives
Welcome (Steve Hedtke)
EPA/NPS objectives in PRIMENet amphibian/UV research (Barbara Walton)
DOI strategy for amphibian monitoring on public lands (Bernard Shanks)
NFS Inventory and Monitoring Program (Gary Williams)
PRIMENet history and objectives (Kathy Tonnessen)
1:50 Overview of workshop planning history and research hypotheses
Workshop planning history and charge to participants (Naomi Detenbeck)
2:00 Plenary presentations
Amphibian monitoring
2:00 . Existing monitoring strategies (Sam Droege)
2:30 Metapopulation approach (Peter Trenham)
UV Dosimetry/Effects
3:00 Early life stage effects/Dosimetry (Gary Ankley/Joe Tietge/Steve Diamond)
3:45 Field study overview (Steve Corn)
4:00-4:15 Break
4:15 - 5:30 Overview of PRIMENet sites (approx. 10 min ea + questions)
Presented by NPS reps (collaboration w BRD researchers where appropriate)
History of amphibian monitoring/species present, background on amphibian habitat, GIS coverages available,
etc.
Tuesday, February 2
Morning: Lg Conf. Room for full group discussions, Lg. conf. room + Council Room for subgroup discussions
8:00 - 9:30 Subgroups: Discussion of research hypotheses (amphibian monitoring, UV dosimetry/effects issues)
9:30-9:45 Break
9:45-10:15 Full group summaries and discussion
10:15 - 11:30 Subgroups: Discussions on integration of hypotheses
11:30 - 12 Full group summaries and discussion
12-1 Lunch
25
-------
Tuesday, February 2
Afternoon
1:00 Discussion of experimental design issues, including site attribute requirements
Two subgroups:
Monitoring designs: populations and UV-dosimetry/effects
Manipulative studies
Wednesday, February 3rd
8 am - 12 noon
Presentations to full group from Tuesday pm subgroups
Discussion
Break-out back to subgroups for working out details in research outline
Afternoon
Lab tours for those interested
26
-------
APPENDIX 2
Hypotheses for discussion:
Working Hypothesis 1:
Testable Hypotheses:
Working Hypothesis 2:
Testable Hypotheses:
Amphibian species are not currently present in habitats where they
would be expected to occur and/or occurred historically, or are
present but in diminished numbers.
The distribution of occupied and unoccupied wetland breeding
habitats conforms to the expectations of metapopulation theory (i.e.,
larger and less isolated wetlands are more commonly occupied).
The fraction of the total wetland breeding habitats that are occupied
is relatively stable over time.
Rates of local extinction are significantly higher in more isolated and
smaller wetland breeding habitats.
Accounting for the effects of wetland isolation and size, exogenous
stressors (e.g., elevated UV, anthropogenic organic compounds,
introduced predators) can be shown to result in significant additional
reductions in habitat occupancy or elevated turnover.
Spatial distribution is a suitable surrogate for temporal trends in
amphibian populations.
Amphibian species exhibit reduced survival and/or malformations at
higher than expected or historic levels
Elevated levels of aquatic UV measured in the field are associated
with higher incidences of malformations.
Early developmental stages of amphibians exhibit greater frequencies
of morphological abnormalities and/or lower survivorship in wetlands
where higher aquatic doses of UV are received.
27
-------
o.
•s
.3
1
•S3
§
.3
rS
c2
J~
1
•o
12
J3
s
o
1
•|^
V
t>
•«
*W Q
T^'g
8.-§
Q C/3
O T3
I|
« 'C
QJ O
Kg
J 0
1 i
1|
V> h
"!'?
OT"«
11
APPENDIX 3
:PRMENet site selection cril
Parks marked with +/- appea
s
2
o
UJ
oo
tJ
O
O
Q
ee!
S
S3
O
Q
<
1
+
+
+
+
+
1 species detectability1
"£
+
+
T
1
tp
1 solar radiation intense?2
+
+
+
Tp
i
"b
| available elevation gradien
+
+
+
+
+
\ available DOC gradients2
+
+
+
+
+
1 vegetation gradients2
'
,
1
+
+
1 easy site access3
+
+
+
'
+
>100 wetlands in park4
+
+
+
+
•>•
+
| metapopulation suitability4
•£
1 wetland habitat mapped?4
+
+
+
+
•
| available historic data4
CS
a> S
'•3 8
feb 1
fi|
a
3
Q
O.
cd
la fa
CD ra
'•3 %
S g
W) £3
.s •§
2 ^
oi
4_T eu
C fa
^ %
C3 CD
OJD S
.3 ;§
.1° 1=1
§§•2
^ 5
.o"E
o.
o ta J5
'S^-i
S s §
ild
SSI
g-^3
"^3 *Q U3
ti§
1^ i
t-s I
all
^ri
.tJ 0.-3 CKfi
343 2 S^
S-S » e
? a OJJ
!> -a en <1>
3 S g, « -g.
i&jit
*5 ex
e ,p
T3 «>
. S «a s s §
Jlflfi
11.311
^ K»
ai *_rH en d>
Sflri
.,.2 -a &
•a f vo
28
-------
APPENDIX 4
Table of amphibian species present in all six PRIMENet sites sending representatives to February 1999 meeting.
Amphibians
Common name
Ensatina salamander
Pacific tree frog
Western toad
Redlegged frog
Tailed frog
Pacific giant sal.
Cascade frog
Northwestern sal.
Long-toed salamander
Rough skinned newt
Olympic salamander
Mtn yellow-legged frog
Blk-bellied slender sal.
Mount Lyell sal.
California newt
Pacific slender sal.
Yosemite toad
Kern Cyn. slender sal.
Columbia spotted frog
Tiger salamanders
Southern leopard frog
Common mudpuppy
Blue Ridge Spring Sal.
Cope's gray treefrog
Jordan's Salamander
Northern Slimy Sal.
E. narrow-mouth toad
S. Appalachian Sal.
S. Red-backed Sal.
Upland chorus frog
Midland Mud Sal.
Black-chinned Red Sal.
Southern Zigzag Sal.
Long-tailed Sal.
Cave Salamander
Eastern hellbender
Santeetlah Dusky Sal.
Junaluska Salamander
Seepage Salamander
Fowler's toad
Breeding
general/
specific
habitat
terrestrial
aquatic/ponds
aquatic/all
aquatic/all
aquatic/stream
aquatic/stream
aquatic/all
aquatic/all
aquatic/pond
aquatic/pond
aquatic/stream
aquatic/all
terrestrial
terrestrial
aquatic/stream
terrestrial
aquatic/pond
terrestrial
aquatic/pond
aquatic/pond
aquatic/pond
aquatic/stream
aquatic/stream
aquatic/ponds
terrestrial
terrestrial
aquatic/pond
terrestrial
terrestrial
aquatic/pond
aquatic/stream
aquatic/stream
terrestrial
egg terr/larv aq
egg terr/larv aq
aquatic/stream
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
aquatic/pond
Discrete Status
habitat
patches?
no
yes •
yes decline?
yes? threatened
no rare
no?
no declining
yes/no
yes
yes
no
no declining
no
no decline?
no
no
yes decline
no
yes threatened
yes
yes
no
no
yes
no
no
yes
no
no
yes
no
no
no
no
yes
yes rare
no?
no
no?
yes
Duration ACA GRS ROM
larval
stage East
<=
16WKS
8WKS
8WKS x
14 WKS
3YRS
2+YRS
14 WKS
12 WKS
12 WKS
20 WKS
3YRS
>1 YRS
?
?
20 WKS
?
12 WKS
?
<1 YR
12 WKS x
<1YR x
5 YRS x
>1YR x
5 WKS x
8 WKS x
8 WKS x
12 WKS? x
8 WKS x
8 WKS x
10 WKS x x
? x
? x
8 WKS x
>1 YRS x
>l YRS x
5 YRS x
2H-YRS x
>1 YRS x
2H-YRS x
8 WKS x
GLA SEK OLY
West
=>
x x
XXX
XXX
X
X X
X
X
X
X X
X
X
X
X
X
X
X
X
X
X
29
-------
Amphibians
Common name
Shovel-nosed Sal.
Green Salamander
Mole salamander
Marbled salamander
Imitator Salamander
Northern cricket frog
Seal Salamander
Pigmy Salamander
Three-lined Salamander
Blue Ridge 2 lined Sal.
Ocoee Salamander
Black-bellied Sal.
Dusky salamander
Eastern gray treefrog
Bullfrog
Spotted Dusky Sal.
Red-backed sal.
Spotted salamander
Four-toed Sal.
Spring peeper
Green frog
Pickerel frog
Red-spotted newt
Wood frog
N. leopard frog
American toad
Breeding
general/
specific
habitat
egg terr/larv aq
terrestrial
aquatic/pond
aquatic/pond
egg terr/larv aq
aquatic/pond
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
egg terr/larv aq
aquatic/ponds
aquatic/all
egg terr/larv aq
terrestrial
aquatic/pond
aquatic/ponds
aquatic/pond
aquatic/pond
aquatic/pond
aquatic/all
aquatic/pond
aquatic/pond
aquatic/pond
Discrete Status
habitat
patches?
no?
yes decline?
yes
yes
no?
yes decline?
no?
no?
no
no
no?
no?
no?
yes
no
no?
no
yes
yes
yes
yes
yes
yes
yes
yes
yes
Duration
larval
stage
2+YRS
12WKS
12WKS
35WKS
2+YRS
12WKS
2+YRS
2+YRS
>1 YRS
>1 YRS
2+YRS
2+YRS
2+YRS
6WKS
>1 YRS
2+YRS
8WKS
12WKS
20WKS
12WKS
>1 YRS
12WKS
12WKS
9WKS
12WKS
8WKS
ACA
East
X
X
X
X
X
X
X
X
X
X
X
X
x-2
x-2
GRS ROM GLA SEK OLY
West
X
X
X
X
X
X
X
X
X
X
X
X
X XX
X
X
X
X
X
X
X
X
X-l X
X
X
30
-------
cs-
J3
!;»
.3
o
o
CL,
W
g
0
g
oo
0
^
O
0
§
C/D
d
q
1
Researcher
x
wetland alteration & amphibians
| Malcolm Hunter
x
NFS inventory including amphibians
1 Cook&Foley
x
<5
terrestrial salamander study/pond breed
| Ken Dodd
x
woodfrog breeding surveys
JimPetranka
x
pesticides levels in wetlands
x
x
amphibian monitoring and inventory
Steve Corn
x
Colorado River wetlands inventory
David Cooper
x
beaver pond inventory
ffl
x
X
characterizing UV in aquatic habitats
•a
w
x
amphibian inventory
LeoMarnell
x
53
w
oo
eg
•o
M
g>
|
•o
1
T3
1
X
amphibian inventory and monitoring
| Gary Fellers
x
yellow-legged frog movements, etc.
$
Kathleen Matthe
x
trout and yellow-legged frogs in the Sie
P?
1 Vance Vredenbu
x
trout and yellow-legged frogs in the Sie
| David Bradford
x
water quality monitoring
Claudette Moore
x
Pacific northwest amphibian surveys
Bruce Bury
x
inventory effort FY2000
00
OH
X
1
1
3
P.,
C3
.a
•3
.S
c
o
o
1 EdHeithmar
x
UV effects on amphibians
GaryAnkley
x
UV effects on amphibians
| Joe Tietge
X
UV dosimetry and effects research
| Steve Diamond
x
amphibian metapopulation applications
| PeteTrenham
31
t.S. GOVERNMENT PRINTING OFFICE: 1999 - SSO-101/20001
-------
-------
-------
m
I
o
o
33
CO
CO
o
CO
en
E.
3-g"
-" co
TJ 3'
3- CD
< co
S. M
CD
o o
5" ®
o 2.
5" CD
JU
1
-j- g BJ_
cn
M -,
o> 3
c» CD
I
CD
ff)
a
m c
< -•
^' CD
O Q.
I 22
§ 8?
3- CD
03 CO
9 -v
T)
CD
a
o
cl
CD
I
s
o
P
-a
m
3J
5m-
PRESORT
OSTAG
E
D
&
P
Q
cb
01
------- |