Effects of the Aquatic Environment on Growth of the Amphibian Pathogen Bactrachochytrium dendrobatidis

Part of the Young Naturalist Awards Curriculum Collection.

by Scott, Grade 11, Arizonia - 2010 YNA Winner


Chytridiomycosis, a disease caused by the fungus Batrachochytrium dendrobatidis, is threatening amphibians worldwide. This research evaluates how the aquatic environment of an amphibian influences the growth of B. dendrobatidis. Water samples collected from amphibian habitats represented diverse geographic conditions which impact the water chemistry of a site. Bioassay experimentation measured optical density over time, indicating the fungus' growth within each water sample. Ion Chromatography (IC) and Ion-Coupled Plasma Spectrometry (ICP) documented the water chemistry. A stepwise multiple regression analysis ([alpha symbol]=0.1) isolated the factors significant to B. dendrobatidis growth, identifying the elements as inhibitors or facilitators. Concluding, the aquatic environment affects B. dendobatidis: growth varied dependant upon the water chemistry of the site (R2=99.91%) and geographic trends were observed. This study is a useful guide for conservation efforts, interpreting how a habitat's water chemistry may protect or predispose amphibians to infection from B. dendrobatidis by affecting the pathogen's growth and potential infectivity.


Despite the fact that amphibians have thrived on Earth for more than 360 million years, through two major Ice Ages and four global warming periods, The Global Amphibian Assessment 2008 reports that 32.4% of the world's amphibians are now known to be threatened or extinct (Gewin 2008, IUCN Red List 2008). Issues contributing to amphibian population decline include habitat destruction by deforestation and human encroachment, pollution, commercial trade, and global warming (Daszak 1999). But Chytridiomycosis, a disease caused by the fungus Batrachochytrium dendrobatidis, is causing catastrophic amphibian die-offs, and was described at the Amphibian Conservation Summit in 2005 as "the worst infectious disease ever recorded among vertebrates in terms of the number of species impacted, and its propensity to drive them to extinction."

When I was a boy growing up in Michigan, I spent many lazy afternoons sitting on the bank of a creek near my home trying to catch frogs. This early scientific exploration held my fascination for hours. When I learned that amphibians are facing the threat of mass extinction, I couldn't accept the thought that one day my children or grandchildren might not be able to enjoy the experiences I remember so fondly. That launched my interest in this research.

The death rate from Chytridiomycosis is high, occurring within 18 to 70 days after exposure. The fungus responsible for this disease, Batrachochytrium dendrobatidis, is an aquatic organism with two life stages: the zoospore, which is the waterborne motile stage, and the zoosporangium, which is the spherical reproductive stage. As the fungus matures, zoospores form within the zoosporangium's cell walls, and are released by a discharge tube that dissolves when wet, making an aquatic environment necessary (Berger 2007).

Many factors affect the chemical composition of an aquatic environment. Chemical contaminants may occur naturally, carried along in groundwater from rocks and sediments, or from land erosion into surface water. They can be human-induced, from industrial discharges or runoff from urban and recreational activities, agriculture, and waste disposal (USGS 2009, Biello 2008). Local conditions also impact water chemistry, including the types of plants surrounding a habitat, the animals that frequent the water source, riverbed composition, water flow, remoteness, and climate.

In my review of the literature, the impact of water chemistry on amphibians is well studied, but study of its effect on the growth of B. dendrobatidis is limited. Research has focused on the Chytridiomycosis infection (Voyles 2009, Daszak 1999, Berger 2007) or treatment approaches (PhysOrg 2007). Jeffrey Piotrowski (2004) and Megan Johnson (2003) studied the fungus' response to environmental stressors in a lab setting, researching pH, UV light, heat, chemical disinfectants, and desiccation. My project uniquely evaluates the effects of inorganic water chemistry on B. dendrobatidis growth, as I hope to describe how the chemical composition of the aquatic environment may be contributing to this pathogen's role in amphibian decline.


Scott at home in an amphibian habitat
Scott at home in an amphibian habitat


In the preliminary phase of my research, I collected water samples from potential amphibian habitats around Phoenix, Arizona, in order to perform a bioassay of B. dendrobatidis to evaluate its growth within the water samples. The results supported my hypothesis that water chemistry in the natural environment affects the growth of B. dendrobatidis. Site pH and temperature were also evaluated, but these factors did not significantly affect fungus growth.

Encouraged by the initial data, I wondered which chemicals in the water directly affected the fungus. This question fueled my continued research, as I predicted that water chemistry was a clue to assessing which amphibian habitats are more or less likely to foster the growth of B. dendrobatidis.

The research question for this project is: Does the water chemistry of an aquatic environment influence the growth of Batrachochytrium dendrobatidis? The overall research hypothesis states that the growth of B. dendrobatidis is dependant upon the chemical properties unique to its aquatic environment. The working hypotheses are that the chemical composition of a water source varies with its geographic location, and that the growth of B. dendrobatidis will vary with the chemical composition of that water source.

The water chemistry (independent variable) is represented by water samples collected throughout Arizona. The dependent variable, growth of B. dendrobatidis, is measured as a function of optical density (% absorbance) over time, reflecting the relative number of zoospores and sporangia produced, and indicating the growth of the fungus (Johnson 2003). The bioassay experimentation was conducted in controlled conditions: the laboratory temperatures (approximately 18˚ to 23˚C) were within the optimal range (Piotrowski 2004), and a control of distilled water was used. There were 55 replicates for each of the 17 water sites, performed across 20 trials, for a total of 935 replicates in the study. Chemical analyses described the water chemistry: ion chromatography (IC) quantified the anions present, and inductively coupled plasma-optical emission spectrometry (ICP) quantified the cations. This data was used to identify potential inhibitors or facilitators of B. dendrobatidis growth, using multiple regression analysis techniques to correlate the chemical composition of the water samples to the growth on the comparative bioassay scans.


Field Observation

Figure 1: Frog believed to be infected displaying rigidity and lethargy.
Figure 1: Frog believed to be infected displaying rigidity and lethargy.

Amphibians include three taxonomic orders, with Anura (frogs and toads) being the most affected by B. dendrobatidis and comprising 31.6% of threatened or extinct amphibians (IUCN Red List 2008). My research doesn't work with frogs directly, but I wanted to understand their plight in Arizona. I met with Glenn Frederick, a district wildlife biologist, and spent a day observing Chiricahua leopard frogs in their habitats, finding several sites in the Huachuca Mountains suspected of being infected by B. dendrobatidis. The fungus' motile zoospores invade the frog's skin and damage the keratin, possibly interrupting its electrolyte balance and leading to cardiac arrest (Daszak 1999, Voyles, 2009). Clinical signs include sloughing of the skin, abnormal posture or rigidity, lethargy, and hemorrhages in the skin, muscle, or eye (Daszak 1999). I observed many frogs that presented as ill, and some that were dead (Figures 1, 2).

Figure 2: Dead frog with sloughing and hemorrhaged skin, tested positive for Bd.
Figure 2: Dead frog with sloughing and hemorrhaged skin, tested positive for Bd.

Water Sample Collection and Analysis

Seventeen water sample sites were selected to represent the major watersheds in Arizona. The sites are documented amphibian habitats (Holycross 2008, Rosen 2009).

Water samples were collected in I-Chem Certified ® high-density polyethylene 125 mL bottles, which are pre-cleaned according to Environmental Protection Agency standards and certified for metals analysis and water-quality testing.

Three samples were collected approximately two inches below the water surface by sterile syringe, and filtered through a 0.22 µm cartridge into the sterile I-Chem bottle. Samples were labeled with site and date: one delivered to the Goldwater Environmental Lab at Arizona State University for chemical analyses (IC and ICP); the other two were used for bioassay testing. Samples were refrigerated.


Figure 3: Ultraviolet Biohazard Flow Hood.
Figure 3: Ultraviolet Biohazard Flow Hood.

Bioassay experiments were conducted in the lab of Dr. Elizabeth Davidson at Arizona State University, which is BSL-2 certified. Experiments were performed under a biohazard laminar flow hood (Figure 3). Procedures as per Johnson (2007) were observed for disinfecting the flow hood before and after each experimental session, including using a misting bottle with a 70% ethanol solution to spray all surfaces inside the hood, as well as autoclave procedures.

Cultivation of B. dendrobatidis

The strain Pseudacris triseriata was used for this project, cultured from a native Arizona frog at Arizona State University. The culture was stored refrigerated in liquid TGhL (tryptone gelatin hydrolysate lactose) media. To allow ample growing room, the culture was passed into a new culture flask every two weeks, maintaining a sterile culture.

For culture growth, plates were prepared as follows: fill petri dish two-thirds full with melted TGhL agar and allow to solidify; extract 1 mL of B. dendrobatidis culture by sterile pipette and completely cover the agar; repeat for each plate being set; seal plates in a Ziploc bag for two days. Cultures were examined under an inverted microscope (Figure 4) to ensure that no bacterial growth was present and that there was sufficient growth to proceed.

Experimental Solutions

Figure 4: Inverted Phase Microscope and 96-well Microtiter Plate.
Figure 4: Inverted Phase Microscope and 96-well Microtiter Plate.

B. dendrobatidis zoospores were isolated as follows: flood each culture plate with 1 mL of liquid TGhL and let stand for 15-20 minutes; extract 1 mL of solution from each plate and filter through a 20 micron nylon filter (to remove most of the sporangia) into a sterile 50 mL tube, assuring all of the solution seeps through. The number of zoospores in solution was estimated using a hemocytometer and an upright microscope.

Each water sample was filtered by drawing 10 mL of the sample into a sterile syringe and passing it through a 0.22 µm sterile filter into a sterile 15 mL tube. Tubes were labeled with site number and date.

Bioassay Experimentation

Experimental solutions were prepared for each water sample using a sterile micropipette to combine 50 µL of the filtered water sample being tested, 50 µL of liquid TGhL, and 100 µL of zoospore solution into a 0.6 mL sterile tube.

A sterile 96-well microtiter plate (Figure 4) was prepared by transferring each experimental solution into a separate well in the plate, properly recording their positions on a template. A sterile plastic cover was placed on the plate to avoid contamination. Each well represents a replicate, and each plate a separate trial.

The optical density (% absorbance) of experimental solutions in each well was recorded using a microtiter plate reader, dated and labeled as Day 0. Scans were recorded again at Days 2, 4, 6, and 8 to document growth of B. dendrobatidis. Results were compared to a distilled water control.

Ongoing qualitative observation of the fungus by inverted microscopy monitored the samples to assess general condition and signs of contamination.


Figure 5: Map of Arizona Watersheds Showing Sampling Site Locations.
Figure 5: Map of Arizona Watersheds Showing Sampling Site Locations.
large  Figure 7: Overall Percent Change in Growth ofB. dendrobatidis for Days 0-8.
large Figure 7: Overall Percent Change in Growth ofB. dendrobatidis for Days 0-8.
Table 2: Average Percent Change in Growth of B. dendrobatidis Days 0-8 per Watershed
Table 2: Average Percent Change in Growth of B. dendrobatidis Days 0-8 per Watershed
FIGURE 8. Average Percent Change in Growth of B. dendrobatidis per Watershed in Arizona.
FIGURE 8. Average Percent Change in Growth of B. dendrobatidis per Watershed in Arizona.
FIGURE 8. Average Percent Change in Growth of B. dendrobatidis per Watershed in Arizona.
FIGURE 8. Average Percent Change in Growth of B. dendrobatidis per Watershed in Arizona.
FIGURE 9. Summary of Multiple Regression Analysis with Stepwise Selection Procedures.
FIGURE 9. Summary of Multiple Regression Analysis with Stepwise Selection Procedures.

Water sample sites are documented in Figure 5. The study location is significant because Arizona is a desert state with limited aquatic habitats, and B. dendrobatidis has been implicated in the decline of native species (Bradley 2002, Schlaepfer 2007). Samples were collected within a three-month period in fall and early winter, after the monsoon season so there is sufficient water flow for collection but before snowmelt runoff. A study of native Arizona frogs by Greogory Bradley (2002) reported an increase in the number of die-offs from Chytridiomycosis during October through February.

Figure 6: Optical Density (% of absorbance) of B. dendrobatidis as a Function of Time (days) Representing the Average Data of all Replicates.


Data from the optical density scans was calculated as the average growth for all replicates corresponding to each of the 17 water sites (Table 1). Figure 6 correlates this data (% absorbance) against time (Days 0 to 8) for each site, to display growth trends. To better represent growth data, the overall percentage change in growth for the eight-day experimental period was calculated for each site and graphically portrayed (Figure 7). The effect of site water chemistry on fungus growth is compared by stratifying Figure 7 into equal segments: the high-growth group (Sites 1, 2, 3, 4, 5) displayed between 160% and 200% growth over the eight days, with Site 1 being the most facilitated; the second strata (Sites 7, 8, 11, 13, 14, 16, 17) demonstrated 120% to 159% growth, representing the majority of the sites; the low-growth group (Sites 6, 9, 10, 12, 15) demonstrated 80% to 119% growth, with Site 12 demonstrating the least growth of all sites. The water control is expected to mark the low end of growth due to its lack of a food source, and this trend was observed. Remarkably, Site 12 grew less than the water control, and Site 6 just slightly more, demonstrating that fungus growth for these water sources was remarkably inhibited.

The water sampling sites represent varied geographic locations, and the water chemistry of the samples is diverse, as noted in the IC/ICP data. Table 2 presents the overall percentage change in growth data as an average for corresponding sites in each watershed. Figure 8 is a map of the watersheds sampled, with Table 2 data presented as a gradient, revealing that growth patterns did vary with geographic location. A chi-square goodness of fit was run on the growth data, and the result was highly conclusive, with a p-value of nearly zero, demonstrating that differences in growth between the sites was not due to sampling variability and can be attributed to water chemistry.

IC/ICP analyses described the water chemistry for each site. Using this data correlated against the average percentage change in growth of B. dendrobatidis (Figure 7), statistical analysis determined which elements are potentially responsible for the variations in growth observed between sites. A multiple regression analysis correlated every factor (element) scanned to the growth observed using a stepwise model, which determines the individual significance and correlation of each factor, as well as the significance of the model as a whole, in order to isolate individual factors. Results displayed in Figure 9 were highly significant. In summary, there were 11 elements that affected B. dendrobatidis growth: sulfate, selenium, barium, and magnesium inhibited growth, whereas bismuth, copper, chloride, thorium, nitrate, antimony, and manganese facilitated growth. These elements accounted for 99.91% of the variation in the total percentage change in growth. Analyzing the resulting t-values for each individual element, bismuth is the most significant factor and antimony the least, although all are well below the set alpha level of 10%, indicating that these elements are statistically significant. Such strong correlation suggests that the null hypothesis should be rejected, meaning that there is statistically significant evidence that growth is dependant on the chemical composition of the source water, and that the regression model is significant.


This project's experimental design was effective in collecting data to understand the correlation between water chemistry and B. dendrobatidis growth. Results strongly support the hypotheses and answer the research question. The research documents that the water chemistry at amphibian habitats throughout Arizona does affect B. dendrobatidis growth, with five sites shown to be highly inhibitory and five highly facilitative (Figure 7). Growth varied depending on the water chemistry of the habitat that was sampled, with a calculated p-value of nearly zero. The elements identified as statistically significant to growth (Figure 9), acting as either inhibitors or facilitators, are likely to be found in the environment. Arizona is a leading copper-producing state, with sulfur as a mining byproduct. Nitrates, ammonia, and magnesium are often found in aquatic ecosystems due to the widespread use of fertilizers (Biello 2008, Robinson 2003), and there is a significant amount of agriculture in Arizona. When the growth data was cross-referenced to the IC/ICP data, Sites 6 and 15 were shown to have high sulfate levels, and Sites 9, 10 and 12 had high selenium levels, all with subsequently low growth, as expected by the model. Locations with facilitated growth rates, Sites 1, 3, 5, 14, and 16, all possessed high amounts of bismuth, with Site 4 containing high levels of copper.

Geographic trends were also observed (Figure 8). Sites with the most facilitated growth were located in the lower Colorado River and lower Gila River basins, representing the western/southwestern part of the state. It is interesting to note that the Verde, Salt, and Upper Gila river basins empty into the lower Colorado and lower Gila River basins, and due to this flow, the buildup of nutrients in downstream basins may support increased growth. Sites with the most inhibited growth were from the upper Colorado and Little Colorado river basins, tryptone gelatin hydrolysate lactose comprising the northern/northeastern part of the state. With the White Mountain Apache Reservation and copper mining operations nearby, agricultural and industrial effects may contribute to the water chemistry. Erosion may also be a factor, as this region suffered a major forest fire, the Rodeo-Chediski, in 2002; Sites 6 and 12 are located in this area. It may also be a matter of geography, as the Mogollon Rim, with an elevation change of up to 2,000 feet, separates these watersheds from the southern half of the state. Varied growth trends may be due to the highly different ecosystems found on the northern versus southern side of the rim. The remaining watersheds were in the middle strata, with no observable trend.

Further research could expand water sampling to assess regional trends, evaluate the effects of organic water chemistry on growth, or compare different strains of B. dendrobatidis. Contamination did occur, decreasing the overall number of replicates in the study, although it is unlikely that the data would change significantly due to the large sample size and the strong correlation demonstrated.


  1. The water chemistry of amphibian habitats representing diverse watersheds throughout Arizona affects the growth of B. dendrobatidis; geographic trends were observed.
  2. Statistically significant variations in the growth of B. dendrobatidis are observed depending on the chemical composition particular to the aquatic environment that was sampled.
  3. There is statistically significant evidence that particular chemical elements affect B. dendrobatidis growth, with some elements acting as inhibitors and others as facilitators.


Amphibians enhance our world in many ways. They are a source for "biomedicines" like analgesics and antibiotics. Amphibians are a critical link in the food web because they prey on insects, which benefits agriculture and minimizes the spread of disease; and they provide a food source for other species, a cycle which is interrupted by population decline, thereby impacting biodiversity (Zippel 2009). Also, amphibians are easily affected by environmental stressors, so they are excellent indicators of overall ecological health.

Results from this research could be used by the Arizona Game & Fish Department to correlate with amphibian die-offs occurring in the state. Since Chytridiomycosis is a global problem, trends observed in Arizona may be helpful to other regions as well. This work may be used as a guide to promote amphibian conservation. Research findings may be helpful to set up and maintain the water chemistry in zoo exhibits or other protected habitats that inhibits B. dendrobatidis growth. Further research could monitor field sites for the water chemistry features that facilitate fungal growth, thereby identifying the amphibians at greater risk for Chytridiomycosis in order to relocate them to safer areas.

Many scientists feel that "saving amphibians is the biggest conservation challenge in the history of humanity" (Gewin 2008). Understanding how water chemistry affects the growth of the amphibian pathogen B. dendrobatidis is critical to identifying methods to control or eradicate it. This project contributes to understanding that link, and I hope it helps to save these wonderful creatures from extinction for future generations to enjoy.


Amphibian Conservation Action Plan. Amphibian Conservation Summit 2005, Washington D.C. Retrieved from the World Wide Web on 17 Jan 2009. http://www.globalamphibians.org/acap_5fsummit_5fdeclaration.pdf

An Analysis of Amphibians on the 2008 IUCN Red List. International Union for Conservation of Nature (IUCN), Conservation International and NatureServe. Retrieved from the World Wide Web on 15 Jan 2009. http://www.iucnredlist.org/amphibians

"Bacteria Show Promise in Fending Off Global Amphibian Killer." PhysOrg.com. Retrieved from the World Wide Web on 22 Sep 2008. http://www.physorg.com/news99134333.html

Berger, Lee, et al. "Life Cycle Stages of the Amphibian Chytrid Batrachochytrium dendrobatidis." Diseases of Aquatic Organisms 68 (2005): 51-63.

Biello, David. "Fertilizer Runoff Overwhelms Streams and Rivers." Scientific American (14 Mar 2008). Retrieved from the World Wide Web on 4 May 2009. http://www.scientificamerican.com/article.cfm?is=fertilizer-runoff-overwhelms-streams

Bradley, Gregory A., et al. "Chytridiomycosis in Native Arizona Frogs." Journal of Wildlife Disease 38.1 (2002): 206-212.

Daszak, Peter, et al. "Emerging Infectious Diseases and Amphibian Population Declines." Emerging Infectious Diseases 5.6 (1999): 735-741.

Gewin, Virginia. "Riders of a Modern-Day Ark." PLoS Biology 6.1 (29 Jan 2008). Retrieved from the World Wide Web on 19 Feb 2008. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2214819

Ground Water Quality. U.S. Geological Survey. Retrieved from the World Wide Web on 4 May 2009. http://ga.water.usgs.gov/edu/earthgwquality.html  

Holycross, Andrew. Interviewed by Scott Boisvert. Arizona State University. 15 Sep 2008.

Johnson, Megan. "Working with Batrachochytrium dendrobatidis." James Cook University. Retrieved from the World Wide Web on 9 Jan 2008. http://www.jcu.edu.au/school/phtm/PHTM/frogs/protocol/bd-protocols.pdf

Johnson, Megan L., and Richard Speare. "Survival of Batrachochytrium dendrobatidis in Water: Quarantine and Disease Control Implications." Emerging Infectious Diseases 9.8 (2003): 922-925.

Johnson, Megan L., et al. "Fungicidal Effects of Chemical Disinfectants, UV Light, Desiccation and Heat on the Amphibian Chytrid Batrachochytrium dendrobatidis." Diseases of Aquatic Organisms 57 (2003): 255-260.

Piotrowski, Jeffrey, Seanna Annis, and Joyce Longcore. "Physiology of Batrachochytrium dendrobatidis, a Chytrid Pathogen of Amphibians." Mycologia 96.1 (2004): 9-15.

Robinson, James L. Water Resources Investigations Report 2003-4182. U.S. Geological Survey. Retrieved from the World Wide Web on 4 May 2009. http://pubs.usgs.gov/wri/wri034182/

Rosen, Phil. Interviewed by Scott Boisvert. University of Arizona. 8 Jan 2009.

Schlaepfer, Martin A., et al. "High Prevalence of Batrachochytrium dendrobatidis in Wild Populations of Lowland Leopard Frogs Rana yavapaiensis in Arizona." EcoHealth 4 (2007): 421-427.

Voyles, Jamie, et al. "Pathogenesis of Chytridiomycosis, a Cause of Catastrophic Amphibian Declines." Science 326 (2009): 582-585.

Weldon, Che, et al. "Origin of the Amphibian Chytrid Fungus." Emerging Infectious Diseases 10.12 (2004): 2100-2105.

Zippel, Kevin. "Why Do We Need an Amphibian Ark?" ActionBioscience (May 2007). Retrieved from the World Wide Web on 27 Jan 2009. http://actionbioscience.org/biodiversity/zippel.html