Here we provide further details of the ongoing experimental tests of mitigation strategies and of the mathematical model presented in Figure 6. Case studies include: (1) treating individuals, (2) treating pond habitats with fungicides, (3) treating pond habitats by drying, (4) reintroductions with disease monitoring, and (5) biocontrol with microcrustaceans.
Ongoing case studies
1. Treatment and release (Alytes obstetricans, Spain and Switzerland)
In the Peñalara Natural Park (Madrid, central Spain), the first known chytridiomycosis outbreak in Europe rendered the population of Alytes obstetricans close to extinction . Tadpole abundance dropped remarkably in successive years (e.g. from more than 5000 to 20 in the pond holding the largest population), considerably increasing the value of each tadpole. Dead or sick adults have never been found in the area, while thousands of dead or dying metamorphs could be easily found. Thus, experimental treatments were restricted to tadpoles.
Pilot tests by J. Bosch (National Museum of Natural Science in Madrid, Spain) verified that tadpoles infected with Batrachochytrium dendrobatidis (Bd) survived after metamorphosis when kept in captivity at more than 21°C, a temperature higher than ambient field conditions. Therefore, every single tadpole found in the area was collected and kept in the laboratory at high temperature. Metamorphosed animals were then released, even though some of them tested positive for Bd by qPCR at the time of release. Prior to release, intensive surveys yielded no metamorphs in the wild. In 2009, the thermal treatment was replaced with itraconazole baths , and detailed studies on infection status and survival of released animals are now in progress. Re-infection of treated animals is possible for both kinds of treatments, and it is too early to recommend its use given the risk of Bd acquiring resistance to itraconazole.
With the possibility of complete extirpation of A. obstetricans in Peñalara Natural Park following outbreaks of chytridiomycosis, a captive-breeding program was established in 2008 by the local government of Madrid, the Museum of Natural History of Madrid (CSIC) and the Durrell Wildlife Conservation Trust. Although the main objective of the program is to maintain a captive breeding colony in case of extinction in the wild, the colony also provides a source for reintroductions. Animals were extremely scarce such that finding founders was difficult. However, because founders were captured from relict metapopulations after 10 years of successive mass mortalities, natural selection has likely occurred. Specifically, it is possible that survivors carry genes fixed by natural selection that confer tolerance (better able to reduce the consequences of infection) rather than resistance (better able to resist Bd infection ). When offspring are produced, decisions will have to be made about selectively breeding for tolerance or resistance. Presently, we have an incomplete understanding about which components of the host response lead to prevention of infection, elimination of Bd, or resolution of disease.
A similar approach is being followed by C. Geiger and B.R. Schmidt (University of Zurich, Switzerland). They collected over-wintered A. obstetricans tadpoles from several ponds. The ponds were relatively simple, man-made ponds with amphibian communities consisting of two newt and two anuran species in addition to Alytes obstetricans. The tadpoles were taken to the laboratory and treated against Bd using itraconazole . A. obstetricans tadpoles were selected as a model to test mitigation strategies against Bd because they are thought to be a significant Bd reservoir. Other amphibian species in the ponds were not treated. Previous mesocosm experiments showed that very few A. obstetricans tadpoles treated with itraconazole became re-infected. This was the case even when they were put back into mesocosms where infected conspecifics were present (C. Geiger and B.R. Schmidt, unpublished data).
2. Pond-level treatments - fungicides (Switzerland)
At the University of Zurich, Switzerland, C. Geiger and B.R. Schmidt conducted mesocosm experiments in which they tested whether pond-level treatments against Bd are feasible. The use of fungicides is a common method to control fungal pathogens in medicine and agriculture, but Kilpatrick et al.  described the use of antifungal compounds in natural wetlands to combat Bd as "radical". Nevertheless, methods developed in aquaculture may be particularly useful for the development of methods to control Bd in natural ponds through the use of fungicides.
Laboratory experiments showed that commonly employed anti-fungal chemicals used in aquaculture and by fish hobbyists can clear Bd infection in tadpoles of the midwife toad Alytes obstetricans (C. Geiger and B.R. Schmidt, unpublished data). Mesocosm experiments were used to learn whether antifungal chemicals are also effective at eliminating Bd from experimental tadpole communities under more natural conditions and how they affect the pond ecosystem. For example, fungi are important decomposers and primary producers in pond food webs  and we need to know how the use of fungicides affect ecosystem functions and services. While direct effects of fungicides on pond organisms may be negligible, indirect effects may have strong negative effects . Even if Bd could be eliminated from natural environments using fungicides, environmental regulations may prevent the use of fungicides in wetlands.
3. Pond-level treatments - drying (Alytes muletensis, Spain)
Drying the habitat containing pathogens can reduce disease incidence. Kriger and Hero  showed that Bd occurs primarily in permanent ponds but was absent from ephemeral ponds. Thus, draining ponds may be a way to suppress Bd in the environment. Because many amphibian species are adapted to ephemeral habitats, draining ponds may not affect amphibian populations negatively , especially when done late in the season when tadpoles are no longer present. If the timing of pond drying can be managed, temporary natural or constructed ponds offer a feasible option for managing amphibians impacted by disease. The construction of temporary ponds is already advocated as an amphibian conservation strategy in highly urbanized areas in Europe  and is used as a remediation measure in the United States.
Midwife toads (genus Alytes) are probably the most Bd susceptible species in Europe [35, 36, 206]. Tadpoles have long lifespans (often more than one year), allowing them to be in permanent contact with zoospores. Adults are highly terrestrial and only males approach the water to release egg clutches. Infected populations of Mallorcan midwife toads, Alytes muletensis, seem to be appropriate targets to explore mitigation approaches for several reasons. Populations are contained in a very dry environment which forces animals to move along torrents and impedes migration among different basins . In addition, no other amphibian species co-occur with A. muletensis, and pools holding tadpoles are small and relatively free of organic material.
The first attempt to eliminate Bd from an A. muletensis infected population is in progress . In this attempt, both individuals and the habitat are being treated. The target pool consists of two small cisterns created for watering live-stock in a short torrent gorge. The Mallorcan midwife toad is the only amphibian species inhabiting the pool, and the scarcity of aquatic vegetation, rocks or mud allow the capture of every tadpole. In several visits during a 6-month period, all tadpoles were collected and taken to the laboratory, where they were treated with itraconazole following Garner et al. . Field work started before the breeding season with collections of over-wintered tadpoles and continued until no new egg-clutches were found and every tadpole was collected. The pool was completely drained. Treated tadpoles were then put back into the pool after the first autumn rains. We expected that the Bd population would not recover once its main tadpole reservoir had been successfully cleared of infection and the ponds dried. However, results from spring of 2010 indicate that reintroduced tadpoles contracted Bd infections, but infections were of a lower intensity. Thus, continued eradication efforts will target adults as well as larvae in the area .
4. Reintroductions of Bufo boreas
Boreal toads have been extirpated from 75% of sites inhabited historically in Rocky Mountain National Park (RMNP) and have declined precipitously in the southern Rocky Mountain Region . In 2007, RMNP launched a thoughtfully planned effort to reintroduce boreal toads. The site was chosen in a region within the park that was historically inhabited by boreal toads . Donor toads were offspring of toads collected in the park and bred in captivity. After three years of surveys, sites were selected based on habitat suitability , proximity to existing toad populations, proximity to human activities, logistical considerations, and disease status. Using molecular methods, disease status was determined from skin swabs of boreal chorus frogs, Pseudacris maculate, and wood frogs, Rana sylvatica and from water samples . Bd was not detected at the selected site. Introductions of tadpoles were initially planned for 3 - 5 years. The project has thus far released tadpoles (700 - 14,000) in three consecutive years (2007-2009), and seven adults in 2008. The adults were excess hatchery individuals and released as an opportunity to assess their usefulness as sentinels for disease. These individuals were monitored using radio telemetry and swabbed weekly between June and September. Diagnostic skin swabs revealed the presence of Bd in ~30% of sentinels, indicating that Bd is still present in the area. In 2009, a handful of one and two year old toads were located at the site. Future releases are planned with extensive monitoring to quantify the success of the reintroduction program.
From this effort, we learned that adults can be effective sentinels for Bd presence, and that continued monitoring is extremely important. Monitoring is often the most neglected part of a translocation project [103, 110, 219] and becomes especially critical when dealing with a transmissible disease and animals that may move relatively long distances. Monitoring is essential and provides information to specify management goals and articulate research hypotheses . As we learn more about the efficacy of mitigation measures, a sound monitoring protocol will be imperative both for disease surveillance and population assessments.
5. Biocontrol with microcrustaceans
The littoral zone of ponds and lakes is a complex area with multiple interactions between biotic and abiotic factors. Especially predator-prey interactions of fish or crayfish and zooplankton, and zooplankton grazing/filter-feeding on smaller microorganisms form a continuous and interdependent cycle over the seasons. The introduction of fish and algae/cyanobacteria into these ecosystems or changes in abiotic factors (pH, chemicals, nutrients, temperature) can destroy established food chains and result in the disappearance or overabundance of some species. The disappearance of cladocerans from Lake Tahoe, CA, in the early 1970's, for example, was linked with high densities of introduced opossum shrimp (Mysis relicta) and kokanee salmon (Onchorynchus nerka) . A recently study performed by Koksvik et al.  also correlated a reduction in cladoceran biomass to the introduction of mysids at Lake Jonsvatn, Norway. Whereas cladocerans were highly affected by introduced fish and shrimp, copepods were not or were less negatively impacted (same study).
As filter-feeding organisms, microcrustaceans can be natural predators of Bd zoospores which are 3-5 μm in size and contain valuable nutrients, especially lipids . The size of food particles successfully filtered depends on the filter apparatus of the microcrustacean species . Furthermore, some microcrustaceans have food preferences and are able to actively select between different food particles . Kagami et al. [185, 226] showed that Daphnia sp. were feeding on zoospores of the chytrid pathogen (Zygorhizidium planktonicum) of the diatom Asterionella formosa.
At the Environmental Studies Area (ESA) at California State University Bakersfield, A. Lauer isolated different microcrustaceans (cladocerans, copepods, and ostracods). These were cultured in freshwater supplemented with 'green water', especially Chlorella sp. as a food source (L.F.S. Cultures, Oxford, MS). The microcrustaceans were incubated at 15°C in an incubator (Percival E-30B) with a 12 h day (40 μmol quanta m-2s-1) and 12 h night cycle and were growing and multiplying well under these conditions. Cultures of Bd (JEL213, obtained from J.E. Longcore, University of Maine) were successfully grown in liquid 1% tryptone medium at 18°C.
In an initial experiment, six Daphnia sp. individuals from ESA were incubated together with ~1.5 × 105Bd zoospores/ml in 5 ml of sterile freshwater (0.22 μm filter sterilized) in wells of a 6-well plate (max. vol. 10 ml) over a period of seven days in comparison to a negative control were no Daphnia were added, and a control were only Daphnia were present, but no Bd zoospores. The experiment was performed in an incubator at 15°C with a day night cycle (12 h under 40 μmol quanta m-2s-1). Prior to the experiment, the Daphnia individuals were rinsed three times with sterile freshwater  to remove transient microorganisms that could be used as a food source. The same experiment was performed with a local ostracod species (seed shrimp, Cypridopsis adusta), isolated from the same environment as Daphnia sp. (pond at ESA at CSUB). The species of Daphnia used in this experiment still needs to be confirmed. Based on microscopic observations it was most probably D. rosea or D. galeata.
After the end of the experiment (day 7), all microcrustaceans that were incubated together with Bd zoospores were still alive at the end of the experiment, whereas all microcrustaceans without zoospores as a food source died. The well with Bd zoospores alone had formed clusters of zoosporangia visible with the naked eye. Microscopic investigations proved that the amount of Bd zoospores was under the detection limit in the wells were microcrustaceans had been present. This experiment demonstrated that the two species of microcrustacean tested were actively feeding on Bd zoospores. Based on this observation, a second exposure experiment was set up where different amounts of Daphnia individuals (0, 4, 8, 12, 16, 20, and 24) were exposed to the same amount of Bd zoospores (~1.6 × 105 spores/ml) over a period of three days in 15 ml of sterile filtered freshwater in sterile collection containers (max. vol. 50 ml) under the conditions described in the pilot experiment. Overall, every 1.5 h a sample of 200 μl was taken which added up to five samples taken each day. The experiment was continued for two more days, taking samples during the day and leaving them untouched during the night. Bd zoospores were stained with Malachite green (a spore stain) and counted using a Phase Contrast Microscope (Nikon, type 104) with 400 × magnification and a Neubauer counting chamber.
Even though the initial number of Bd zoospores/ml were supposed to be ~1.6 × 105/ml, the numbers varied from 2.3 - 4.2 × 105/ml at the beginning of the experiment. This is probably due to clusters of zoospores that were initially counted as one and then broke apart when the Bd-dilutions were prepared. A decline in number of Bd zoospores was observed for all trials with a steeper decline when 12 and more Daphnia were present in comparison to the control where no Daphnia were added to the zoospores (Main text, Figure 5B). After the steep decrease in zoospores on day one of the experiment, the zoospore count stabilized or declined moderately over the next two days (Main text, Figure 5B). It was observed that the Daphnia individuals started reproducing in all containers on day 3 of the experiment. Three counts of zoospores were combined and an average calculated for each sample investigated. The exact amount of Daphnia individuals at the end of the experiment, their size and biomass was not determined.
In an ongoing project, A. Lauer has begun studying the diversity of microcrustaceans and the presence of Bd in different ponds of the Southern San Joaquin Valley and the foothills of the Sierra Nevada (CA), comparing the diversity in the spring and fall. In addition to investigating the microcrustaceans present in the watercolumn, sediment samples were analyzed with molecular methods for the presence of diapausing eggs, which are known to survive for hundreds of years . A discrepancy between the diversity of microcrustaceans in the watercolumn to the one in the sediment might indicate a shift or decline in diversity of microcrustaceans due to environmental influences, such as pollution, eutrophication, invasive species, including the introduction of fish.