The Role of Infectious Disease in Amphibian Population Decline and Extinction James P. Collins School of Life Sciences Arizona State University University of Tennessee Department of Forestry, Wildlife and Fisheries Center for Wildlife Health Knoxville, Tennessee 22 April 2010 OUTLINE Introduction: A grand challenge problem A model system: What is the evidence for ranaviruses as a cause of amphibian decline and extinction? A model system: What is the evidence for the amphibian chytrid fungus as a cause of amphibian decline and extinction? Conclusions Introduction: Two grand challenges for 21 st century environmental biology 1. Global loss of biodiversity 2. Emerging infectious diseases Do emerging infectious diseases have a role in the decline and extinction of species? 1
A theoretical problem When is emerging infectious disease a force in extinction? Conventional theory suggests that a pathogen is unable to drive a host population to extinction (Kermack & McKendrick 1927, Anderson & May 1979) Assumptions: density dependent transmission homogeneous mixing no alternate hosts no environmental reservoirs Proportion of susceptible hosts that become infected DENSITY DEPENDENT TRANSMISSION Density (# infected / volume) Susceptible Transmission Infected Proportion of susceptible hosts that become infected DENSITY INDEPENDENT TRANSMISSION Prevalence (# infected / total # of hosts) Empirical claims Proposed declines due to EIDs Chestnut blight Dutch elm disease Sudden oak death White nose syndrome in bats Coral bleaching disease 2
Empirical claims Proposed major extinctions due to EIDs Hawaiian birds (Reynolds et al. 2003) Amphibians (Stuart et al. 2004) Pleistocene large mammals (MacPhee & Marx 1997) Proposed single species extinctions due to EIDs Polynesian snail (Daszak and Cunningham 1999) Sharp-snouted day frog (Schloegel et al. 2006) Endemic Christmas Island rat (Wyatt et al. 2008) Causes of amphibian declines Commercial use Introduced species Land use change Contaminants Climate change Infectious disease Amphibian diseases: Macroparasites Johnson et al. flatworm parasites and deformities Trematode-agriculture runoff hypothesis (Kiesecker et al. 2002; Johnson et al. 2007; Rohr et al. 2008) Alaria: a trematode parasite of amphibians Pacific tree frog Scan by Sessions & Ballengee 3
Amphibian diseases: Microparasites Bacteria Protozoa Saprolegnia ferax Chytrid a fungal pathogen of frogs and salamanders - Batrachochytrium dendrobatidis (Bd) Ranavirus a genus of viruses infecting cold blooded vertebrates A model system: Ranavirus Do the transmission dynamics of Ambystoma tigrinum virus (ATV) place ATV-infected amphibian populations at risk for pathogen-induced extinction? Amphibian diseases: Viruses Viral groups that infect amphibians: 1. Adenoviruses 2. Caliciviruses 3. Flaviviruses 4. Parvoviruses 5. Retroviruses 6. Togaviruses 7. Herpesviruses Pathogenic 8. Iridoviruses Herpesviruses - cause rare renal tumors in frogs; not implicated in decline or extinction Iridoviruses - common and involved in epidemics 4
Amphibian diseases: Iridoviruses Genera of Iridoviridae: 1. Iridovirus infects invertebrates, mainly insects 2. Chloriridovirus infects mosquitoes 3. Lymphocystivirus infects fish 4. Ranavirus infects salamanders, frogs, fish, and reptiles Amphibian diseases: Iridoviruses Eight Ranavirus strains are reported from amphibians and may infect multiple species or only one: Bohle iridovirus infects amphibians, fish, and reptiles. Frog virus 3 is reported from almost a dozen frog species and one salamander species. Ambystoma tigrinum virus is reported only from salamanders. Ranavirus relationships 100 94 100 100 100 FV3 Frog 100 100 99 100 SSTIV TFV ATV EHNV GIV SGIV LCDV-1 Turtle Frog Salamander Fish LCDV-C Fish ISKNV 100 OSGIV 100 RBIV CIV Insect MIV 0.1 (Jancovich et al. 2003. Virology) 5
Ranavirus host range and origin? Ranavirus genomic DNA analysis Suggests that the most recent common ancestor of Ranavirus was a fish virus followed by a jump from fish to salamanders or frogs Human involvement in Ranavirus host shifts? Movement of hosts/disease by humans? Movement of tiger salamanders as bait (Source: Picco and Collins. 2008. Conservation Biology) 6
Amphibian commerce as a source of pathogen pollution Wild populations Wild populations, lakes virus? (Source: Picco and Collins. 2008. Conservation Biology) ATV and population dynamics Host: Ambystoma tigrinum nebulosum Pathogen: Ambystoma tigrinum virus (ATV) Study area Kaibab Plateau 7
Salamander life history (Brunner et al. 2004) 8
Virus transmission ATV is transmitted by direct physical contact (bumping, biting, and cannibalism) as well as by necrophagy and indirectly via water and fomites. Larval salamanders become infectious soon after exposure to ATV and their propensity to infect others increases with time. (Source: Brunner, Schock, Collins. 2007. Transmission dynamics of the amphibian ranavirus, Ambystoma tigrinum virus. Diseases of Aquatic Organisms) Testing density dependent transmission. The experiment varied number of susceptible hosts and number of infected hosts within a 55 L aquarium. (Amy Greer) Susceptible hosts Infected hosts Volume (L) Density (I/V) Prevalence (I/N) Replicates 1 1 55 1 0.5 7 8 8 55 8 0.5 4 40 40 55 40 0.5 4 1 8 55 8 0.89 7 80 8 55 8 0.09 4 1 1 55 1 0.5 3 40 40 55 40 0.5 3 Experimental design: 1059 larvae Infected hosts (I) Sham infected hosts (I c ) 5 days Susceptible hosts (S) Housed individually for 28 days Volume = 55L aquarium Exposure time = 24 hours Laboratory diagnostic testing Housed individually for 28 days 9
Experimental design: 1059 larvae Results 504 susceptible salamanders exposed in treatment replicates 468 developed signs of infection and died 36 uninfected after 28 days (no sub-lethal infections) Proportion of susceptible larvae infected at each density. Vertical bars equal 1 SE. ceptible hosts e infected Proportion of sus that became 1 0.8 0.6 0.4 0.2 0 0 10 20 30 40 Density of infected hosts (# I / 55L) 10
Estimating the transmission constant (β) for a range of transmission functions Type of transmission Function β (units) Additional parameters (units) Neg. loglikelyhood AIC c value ΔAIC Akaike weight q = 0.255 Power βsi q 1.38 (H -q (dimension- day -1 ) less) 20.20 44.9 0.00 0.588 Negative binomial k ln (1+βI/k)S 7.72 (H -1 day -1 ) k = 0.578 (day -1 ) 20.66 45.9 0.92 0.371 Constant risk, asymptotic, density dependent, and frequency dependent functions were evaluated but none was significant. What explains the high transmission rate? Transmission via water? Asymptomatic animals shed high numbers of viral particles per day Non-random contact? Behavioral changes at low densities cause clumping of hosts (Brunkow and Collins 1998) ble hosts infected Proportion of susceptib 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 5 10 15 20 25 30 35 40 45 Density of infected hosts (# I / 55L) What about transmission in the field? 11
Field surveys Mark-recapture Tissue samples for ATV screening Environmental data Results 100 80 Percentage of 60 ATV positive 40 ponds 20 0 p=0.05 p<0.001 <25% 25% - 75% p>0.05 >75% Percent emergent vegetation How does habitat affect transmission? Hypothesis: Habitat fragmentation buffers disease transmission by decreasing larval contact rates Prediction: Contact rate Amount of vegetation 12
Experimental design Sparse vegetation Dense vegetation Mark-recapture by site of origin Tissue sample Larval distribution in a pond varies with amount of emergent vegetation 14% 4% 5% 6% 13% 12% 7% 8% 11% 9% 10% Sparse Dense Results Fragmentation Halo effect A lower incidence of ATV in heavily A lower incidence of ATV in heavily vegetated ponds is caused by lower effective density rather than buffered transmission 13
Host pathogen theory A pathogen is unable to drive a host population to extinction (Kermack & McKendrick 1927, Anderson & May 1979) Assumptions Density dependent transmission Yes Yes Homogeneous mixing No Yes No alternative hosts N/A Yes No environmental reservoirs No Yes Empirical tests of the theory Do the transmission dynamics of Ambystoma tigrinum virus (ATV) place ATV-infected amphibian populations at risk for pathogen- induced extinction? Best evidence suggests the answer is No: Amy Greer et al. (2008) Jesse Brunner et al. (2007) 14
Model system: Amphibian chytrid Do the transmission dynamics of Batrachochytrium debdrobatidis (Bd) place Bdinfected amphibian populations at risk for pathogen-induced extinction? The amphibian chytrid By the mid-1990s it was suspected that the chytrid fungus Batrachochytrium dendrobatidis might be an emerging infectious amphibian disease (EID). EIDs are diseases that are newly recognized, newly appeared in a population, or rapidly increasing in incidence, virulence, or geographic range. Amphibian chytrid life cycle (Source: Rosenblum et al. 2010. PLoS Pathogens) 15
Chytrid - amphibian system Chytrid is associated with anuran declines and extinctions in Australia, Europe, Africa, Central, South, and North America, but also coexists with nondeclining species. It infects most amphibian species tested with effects varying from no clinical disease to 100% mortality. (K. Lips) Microenvironment affects susceptibility to chytrid. p = 1 (habitat can support chytrid) p = 0 (habitat cannot support chytrid) The model was also extended using Eastern Hemisphere data. (Source: Santiago Ron. 2005) Distribution of threatened amphibians in Central America, Northern South America, and the Caribbean (Source: Global Amphibian Assessment 2004) 16
Hypothesis: Amphibian population sizes and species richness decrease as Batrachochytrium spreads 1987-88 Rate of spread ~28 km/yr 1993-94 2002-0303 1996-97 (Source: Lips et al. 2006. Infectious disease and global biodiversity loss: pathogens and enigmatic amphibian extinctions. PNAS) Prediction: Abrupt change in amphibian density and species richness when chytrid arrives Amphibian density or species richness Time Test: estimate change in slope and date of change by fitting a segmented linear model to the data Amphibian density changes along streams chytrid emerges Amphibian densities and segmented linear models for riparian and terrestrial transects (1998 2005) at El Cope, Panama. There was a significant change in slope for riparian (θ2 = -1.36 10-2, t = -24.44, df = 486, P < 0.0001) but not for terrestrial transects (θ2 = -1.74 10-3, t = -0.71, df = 212, P = 0.4802). 17
Amphibian species changes along streams chytrid emerges Amphibian species richness and segmented linear models for riparian and terrestrial transects (1998 2005) at El Cope, Panama. There was a highly significant change in slope for riparian transects (θ2 = -6.45 10-3, t = -6.97, df = 486, P < 0.0001) but not for terrestrial transects (θ2 = -3.77 10-3, t = -1.78, df = 212, P = 0.0757). Central American pattern of declines 50% of ~75 species gone in 4-6 months Remaining species are at 10% of abundance Batrachochytrium likely spread by frog-frog and frog-environment contact Pattern not consistent with land use change, exotic species, commercial use, climate change, or contaminants Conclusion: Batrachochytrium is likely cause of enigmatic declines in this region 1. Extinctions in Costa Rica and Panama affected lowoccupancy and endemic species resulting in homogenization of the remnant amphibian fauna. 2. Extirpations resulted in phylogenetic homogenization at the family level and ecological homogenization of reproductive mode and habitat association. 3. Amphibian declines in this region are an extinction filter, reducing regional amphibian biodiversity to highly similar relict assemblages. 18
What happens after Batrachochytrium emerges? Batrachochytrium in frog populations after decline (Queensland, Australia; 1994-98) Gastric Brooding Frog Declined to extinction in 1985-86 Eungella Day Frog Abundant, then sudden decline in 1985-86 Now persists in a few small populations Eungella Plateau Batrachochytrium was the suspected cause of the declines and is now endemic [Source: Retallick et al. 2004. PLoS Biology] Our results show (i) genotypic differentiation among isolates, (ii) proteomic differentiation among isolates, (iii) no significant differences in susceptibility to caspofungin, (iv) differentiation in growth and phenotypic/morphological characters, and (v) differential virulence in B. bufo. Mass spectrometry has identified a set of candidate genes associated with inter-isolate variation. Our data show that, despite its rapid global emergence, Bd isolates are not identical and differ in several important characters that are linked to virulence. 19
Our results show that cutaneous microbes are a part of amphibians innate immune system, the microbial community structure on frog skins is a determinant of disease outcome and altering microbial interactions on frog skins can prevent a lethal disease outcome. A bioaugmentation strategy may be an effective management tool to control chytridiomycosis in amphibian survival assurance colonies and in nature. A theoretical problem When is emerging infectious disease a force in extinction? Conventional theory suggests that a pathogen is unable to drive a host population to extinction (Kermack & McKendrick 1927, Anderson & May 1979) Assumptions: density dependent transmission homogeneous mixing no alternate hosts no environmental reservoirs A theoretical problem Assumptions: density dependent transmission homogeneous mixing no alternate hosts no environmental reservoirs The amphibian chytrid fungus meets three conditions that could result in a pathogen causing extinction: (1) density-independent transmission (2) non-homogeneous mixing (3) alternate biotic (amphibian) hosts 20
Conclusion: Biodiversity loss and emerging infectious disease Two pathogens associated with enigmatic amphibian declines Chytrid a fungal pathogen of frogs and salamanders - Batrachochytrium dendrobatidis (Bd) Ranavirus a genus of viruses infecting cold blooded vertebrates Proportion of susceptible hosts that become infected DENSITY DEPENDENT TRANSMISSION Density (# infected / volume) PERSISTENCE Proportion of susceptible hosts that become infected DENSITY INDEPENDENT TRANSMISSION Images: A. Hyatt and L. Berger Prevalence (# infected / total # of hosts) DECLINE / EXTINCTION Causes of amphibian declines Commercial use Introduced species Land use change Contaminants Climate change Infectious disease A modern extinction event 21
Sustainability Back-up slides A great disappearing act and its grand challenge questions Back up slides Ranavirus host range and origin 1 3 2 EHNV Grouper-like viruses LCDV-like viruses 22
Ranavirus host range and origin 1 3 2 EHNV Grouper-like viruses LCDV-like viruses 23