The Demise of the Golden Toad and the Creation of a Climate Change Icon Species

Copyright: © Ochoa-Ochoa et al. 2013. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and distribution of the article, provided the original work is cited. Abstract There is an unavoidable degree of uncertainty associated with future climate projections, and even more unpredictability about the potential impact of different climate scenarios on the ecology and distribution of organisms. Conservationists face a major public communications challenge to both raise awareness and mobilise support for conservation and climate change mitigation/adaptation policies while realistically representing complex and uncertain scientifi c information. Here, we illustrate the interplay of these competing communication goals through a review of the representations of the golden toad in the print media and peer-reviewed literature (in English and Spanish). Since its disappearance in 1989 the toad has become an important conservation fl agship species that has been frequently portrayed as the fi rst verifi ed extinction attributable to global warming. Moreover, there was an increase in the certainty of published news items regarding the toad and its demise, especially in the late 1990s. The uncertainty surrounding the toad’s disappearance (apparent in the primary research literature) was poorly represented in the popular press. The transformation of the toad into an iconic species for climate change advocacy may refl ect a perceived need to supply tangible evidence of biodiversity consequences arising from climate change and highlights the challenges facing conservation scientists in communicating scientifi c concerns and uncertainty via the media.


INTRODUCTION
"Perhaps the most common outcome of the scientifi c process is not facts, but uncertainty" Friedman et al. (1999: vii) "For many, false prophecy is still less frightening than uncertainty" Reading (2004: 15) Global concern about the conservation status of amphibians began to gather momentum at the fi rst World Conference of Herpetology in 1989 (Sarkar 1996).Numerous scientists at this meeting argued that there was a general declining trend among amphibian populations in different parts of the world, but with no obvious single cause (Blaustein and Wake 1990;Blaustein 1994;Houlahan et al. 2000).Several hypotheses were suggested, such as ultraviolet radiation, pesticides, introduction of alien species, toxicants, deforestation, and pathogens (reviewed in Blaustein et al. 1998;Collins and Storfer 2003;Cushman 2006; see Appendix 1).These threats are not mutually exclusive and some have proved diffi cult to identify defi nitively or separate in the fi eld (Collins and Storfer 2003), possibly because sub-population decline is inevitable in metapopulation with high demographic variability-a characteristic of many amphibian populations (Alford and Richards 1999;Gillespie 2010).
The golden toad, Incilius (Bufo) periglenes (Savage 1966), is a classic example of a species for which the causes of decline and eventual extinction are poorly understood, but which nevertheless has become an important species within climate change discourse.The golden toad was fi rst discovered in the cloud forest of the Monteverde region of Costa Rica in 1964 (Savage 1966).The small population size and extremely localised geographic distribution meant that the toad was always a conservation concern and, in 1972, a small reserve of less than 4 sq.km was established that encompassed the entire known global population (Crump et al. 1992).Subsequently the reserve was expanded to encompass around 105 sq.km.
Although the golden toad was diffi cult to survey accurately within the dense undergrowth of the cloud forest, accurate counts could be made for a few weeks in April when individuals emerged from the forest to mate in temporary pools (Crump et al. 1992).Thus, unlike the vast majority of extinctions (cf Ladle 2009), the fi nal, rapid disappearance of the golden toad was well documented.More than 1,500 toads were observed in 1987, but only a single toad was observed at the main known breeding site in 1988 and 1989 while seven adult males and two adult females were recorded 4-5 km away in 1988.No verified sightings have been reported since (Crump et al. 1992;Pounds and Crump 1994;Sarkar 1996).However, published accounts of the procedures used to survey and monitor amphibians in the Monteverde cloud forest reserve are limited and rather imprecise.Pounds et al. (1997Pounds et al. ( : 1316) ) write that the "Monteverde reserve has been almost constantly patrolled for 25 years," a comment that Wake and Vredenburg (2008) cited as "daily monitoring."To our knowledge, there is no published account (in print or online) on the frequency of toad monitoring, the extent of the monitored area after 1990, the proportion of the research area that has been systematically sampled, sampling method (day or night samplings), or whether any other forests in the vicinity have been searched.This apparent lack of systematic sampling of the reserve and/or of the wider region introduces a degree of uncertainty into data on the rate and timing of the observed decline (Crump et al. 1992;Sarkar 1996).
Initially, the toad's disappearance was linked to the severe neotropical droughts of 1987-1988 attributed to El Niño-Southern Oscillation (ENSO) conditions (Crump et al. 1992;Pounds and Crump 1994).However, shortly after the toad's disappearance, Crump et al. (1992) acknowledged that other factors (e.g., non-specifi c pathogen attacks) might have played a role in the extinction and also commented that impacts of prior environmental degradation (initiated before monitoring work began) could not entirely be ruled out.The articles based on population monitoring leading up to the crash made no mention of disease playing a role in the toad's decline.Indeed, Crump et al. (1992) commented that data to address this possibility were lacking.Pounds et al. (1999) subsequently published a high-profi le article in Nature arguing that climate change was probably responsible for the decline or disappearance of a number of species of birds, reptiles, and amphibians in the area, citing the golden toad as a specifi c example.Pounds et al. (2006) provided additional climate-trend analyses and developed a more refined argument for the toad's extinction based on the temperature-sensitivity of the behaviour of the pathogenic chytrid fungus Batrachochytrium dendrobatidis (but see Lips et al. 2008;Rohr et al. 2008;Cheng et al. 2011;Garner et al. 2011, for counterarguments against this hypothesis).Pounds et al. stated that their data supported with "very high confi dence" the case for large-scale climate warming being key to the loss of a number of amphibians and implicated this temperature-sensitive chytrid as an integral component of the causal nexus leading to amphibian declines.They focused on species in the genus Atelopus (harlequin frogs) for the most part, but also cited the golden toad as subject to the same drivers (Pounds et al. 2006).Signifi cantly, these later arguments did not cite any new evidence regarding the extinction of the golden toad.
To attribute climate change as the single major cause of any extinction event is problematic (Whitfi eld et al. 2007;Ladle 2009) because there is often no proper way of scientifi cally distinguishing a specific climate-change influence from historical variability in climate and other environmental conditions (Anchukaitis and Evans 2010).The arguments of Pounds et al. (1999Pounds et al. ( , 2006) ) that climate change was the main causal factor in the extinction of the golden toad coincided with a marked increase in climate change discourse in conservation generally.This followed a shift from 1987 to 1992 in the United Nation's focus from "poverty reduction in developed countries" to the "biodiversity crisis," including climate change, biodiversity, and forests (McManus 2000).Climate change was one of the main issues during the United Nations Conference on Environment and Development (UNCED) in Rio 1992, the Conference of the Parties in Geneva 1996 (COP2), and Kyoto 1997 (COP3).
More generally, the focus of the international community on climate change has had signifi cant consequences for conservation (Jepson and Ladle 2010), not least because many sources of conservation research funding have become linked to understandable concerns about how changing climate might impact wildlife and ecosystems.Many conservationists and conservation organisations quickly aligned behind this increasingly dominant environmental theme (Ladle et al. 2005;Ladle and Jepson 2010).In this context, the transformation of the golden toad from an obscure species of mainly herpetological interest into a prominently cited example of a contemporary extinction event (e.g., Pearson 2011) and a global icon for climate change (Stork and Samways 1995) provides an intriguing window into the use of scientifi c information for conservation advocacy.
To better understand the demise of the golden toad and how its story has been used and relayed, we present a comparative assessment of the reporting of its extinction in the academic literature (peer-reviewed journal articles) and in the news media (internet and newspapers).We also reflect on the potential implications for public perceptions of science and conservation.

MATERIALS AND METHODS
We searched the peer-reviewed journal literature via two databases: Scopus and the ISI Web of Knowledge.We obtained six results from Scopus that included "Bufo periglenes" (title, keywords, and body), and six that included "golden toad."From the ISI Web of Knowledge, we obtained 18 articles that included the term "Bufo periglenes" and 29 that included "golden toad" (Appendix 2).We did not fi nd any peer-reviewed article that included "Incilius periglenes."In total, we retrieved 40 peer-reviewed articles published between 1972 and 2010.
We used the LexisNexis ® database to search all news articles from all available sources of published information (newspapers, newswires, magazines, broadcast transcripts, and some blogs) that contained the words "golden toad" and "sapo dorado" (common name in Spanish).This search included articles published and captured electronically from January 1983 to March 2010.In all, we obtained 530 articles published in Spanish and English.After screening and deleting duplicated news or reports not related to amphibians (such as cultural or sports news), we compiled a database of 400 articles.
We recorded the following information for peer-reviewed and news articles: title/headline, month, year, type of article/ news, name of the source, country of release, and whether the author(s) mentioned other amphibian species or other species in general.We categorised presumed causes of disappearance for I. periglenes or population changes for other specifi cally mentioned amphibian species as: climate change, habitat loss (including deforestation), pollution (including pesticides, chemical wastes), ultraviolet radiation, alien species, acid rain, ENSO (i.e., climate variability as opposed to long-term change), and diseases (including viruses and fungi).If the identifi ed cause of population change did not fi t into the above classifi cation, we recorded it as "other causes" and recorded any additional information on population status (decline, disappearance, extinction, etc.) mentioned in the article.
To further identify and characterise reports in media sources and academic journals we searched for signifi er keywords or statements in the headline or in the main body of the text such as: wiped out, extinct(ion), probably or believed to be extinct, not seen, endangered, threatened, disappeared, vanish(ed), declining, killed (fatal), mass extinction and "canaries in a coal mine."We also recorded other information contained in media reports such as direct quotes attributed to non-governmental organisations (NGOs), university researchers, and government offi cials, as well as events such as conferences, and scientifi c articles related to the published news.Additionally, we surveyed 92 peer-reviewed articles (selected haphazardly from 385 papers identifi ed via Scopus in January 2011), which cited Pounds et al. (2006), to specifi cally assess how authors represented this important article in the scientifi c literature.To avoid operator variance the fi rst author performed all literature research, data extraction, and compilation to standardise interpretation and classifi cation of the information.

RESULTS
News items containing the term "golden toad" increased irregularly over time with four peaks in frequency (Figure 1) and an annual mean of 14.86 ± a standard deviation of 15.39.In contrast, the number of peer-reviewed articles in scientifi c journals containing either "golden toad" or "Bufo periglenes" or "B.periglenes" was consistently low over the study period, with a mean of 1.33 ± 1.33 articles per annum.The frequencies of news items and peer-reviewed articles were not correlated (adjusted R 2 = 0.002, P = 0.31, Figure 1).

Figure 1 Number of peer-reviewed articles (sources: ISI Web of Knowledge and Scopus) and news (via LexisNexis®) published
containing the term "golden toad" and/or "Bufo periglenes".The description of the golden toad was in 1966, following which we could not fi nd peer-reviewed articles until 1983.
The fi rst news in the LexisNexis ® database occurred in the same year.
Authors of popular articles invoked different causes for the disappearance of I. periglenes from Monteverde through time.Several causes were identifi ed shortly after the extinction event was reported; those causes typically were described as having high uncertainty.This changed over time and climate change and the amphibian fungal disease, chytridiomycosis, became the dominant reported causes by 2000 (Figure 2).As the number of articles varied through time, we used the proportion of the number of times that different causes were mentioned within the news to illustrate these trends (Figure 2).It should be noted that some of the articles discuss other amphibians in addition to I. periglenes.
We found it more difficult to quantify any trends in statements regarding causes of declines in the peer-reviewed articles due to the low number of published articles containing the search terms (Appendix 2).From the 40 peer-reviewed articles retrieved from the search, we were only able to use 32 in the analysis.One article provided only a description of the toad, three others were specifi cally related to life history, and we were unable to obtain copies of another four.Most remaining articles evaluated more than one cause.A large proportion (17) of the peer-reviewed articles highlighted the golden toad as an example of unexplained decline, another seven mentioned weather variability including ENSO, another fi ve chytridiomycosis, and one the pet trade.Six articles mentioned climate change as the main factor driving the extinction (Appendix 2).The change in the strength of the rhetoric used by the news media to tell the story of the golden toad is also refl ected in the increase in value-laden adjectives such as "wiped out" or "vanished", which peaked in 2006 (Figure 3).The use of the phrase "mass extinction" in the news media also peaked that year.
Whilst a relatively small number of peer-reviewed articles have been published about the golden toad, some have had considerable impact throughout the popular and scientifi c media, most notably, the article by Pounds et al. (2006).This article has been cited for various reasons.Forty-eight articles cited Pounds et al. (2006) as an example of disease impacts promoted by climate change or by climatic or environmental variability.Twenty-one articles cited it as an example of species vulnerability to climate change per se, while six other articles referred to it as an example of chytridiomycosis, and nine challenged the hypothesis presented by Pounds et al. (2006).Finally, eight articles cited the paper as illustrating other causes of amphibian declines (Appendix 3).

DISCUSSION
The observed change in the public media's descriptions of potential causes of the golden toad's extinction and of the uncertainty associated with them could have been infl uenced by the way climate change has come to dominate discussions of global change (Liu et al. 2011).Authors of articles in the public media arguably refl ected some of the tone of high-profi le scientifi c articles and their associated press releases, such as Thomas et al. (2004) and Pounds et al. (2006).As commented on elsewhere, the former article was widely misrepresented in the news media, partly due to misunderstanding of the uncertainties involved in complex modelling studies and partly as a result of extrapolations contained in the associated press release (Ladle et al. 2004(Ladle et al. , 2005)).
The four recent peaks in the frequency of popular news items about the golden toad (Figure 1) roughly coincide with particular Although high-profi le academic articles may generate news stories, we found no quantitative association between the frequencies of the news-media items we surveyed and that of peer-reviewed published articles about the golden toad.This is not surprising given that the global news media does not typically consider peer-reviewed articles newsworthy by themselves, especially those in less prominent journals.Moreover, the publication of peer-reviewed articles is dictated by factors not necessarily in step with the public media's focus at any given time.
Like the Thomas et al. (2004) study, the Pounds et al. (2006) article is a complex modelling study published in a high-profi le journal that suggested more certainty to some readers than often is stated in scientifi c articles.For example, the authors stated in the opening paragraph (which served as an executive summary): "we conclude with 'very high confi dence' (>99%, following the Intergovernmental Panel on Climate Change, IPCC) that large-scale warming is a key factor in the disappearances [of 67% of 110 species of harlequin frogs and the golden toad]."They concluded the abstract with the non sequitur that "the urgency of reducing greenhouse-gas concentrations is now undeniable."Language such as 'very high confi dence' and 'undeniable' conveys technical meaning while also providing strong words for global news stories in the popular media.While the phrase 'very high confi dence' still denotes uncertainty, it indicates a very small measure of doubt, unlikely to produce concerns in the minds of the lay reader.Indeed, considering that the chain of argumentation in the paper relies on several untestable assumptions (given the lack of direct monitoring or other autoecological data and thus the poor resolution of what drove the population declines), the use of such a strong expression of confi dence appears unwarranted, and indicative of an effort to downplay uncertainty.The use of such a phrase can have consequences similar to those arising from the use of another technical expression common in bioclimatic envelope modelling-'committed to extinction' (e.g., Thomas et al. 2004), which means en route to eventual extinction if no mitigation is undertaken, but which frequently has been simplifi ed to 'will be extinct by date x' by the global news media (Ladle et al. 2005;Ladle and Jepson 2010).
It is hard to determine why it is that particular over-simplifi ed narratives encapsulated by statements such as… "golden toad driven to extinction by climate change"

Figure 3 Frequency of use of different labels referring to the disappearance of the golden toad in news items
or "one million species extinct by 2050" gain traction via the media and conservation NGOs, when other explanations or narratives published in similarly respectable or highly cited journals do not (but see for e.g., Ladle et al. 2005).For instance, following the Pounds et al. (2006) paper, the golden toad became a major element of the story in the media, despite the main focus of the article being on a completely different genus, Atelopus.Although the golden toad is mentioned in the abstract alongside Atelopus, the article provides little further mention (and no new data) of the demise of the golden toad.In contrast, a recent article arguing that the extinction of the golden toad was caused by increased climate variability rather than directional climate change (Anchukaitis and Evans 2010) was not reported widely in public media, despite also being published in a high-profi le journal.
The elevation of the golden toad to iconic status illustrates how selected writings in scientifi c journals and the popular media are related and interpreted, despite scientific uncertainties or improved information over time.The biggest NGOs were quick to use the rhetorical potential of the golden toad and have been instrumental in turning the toad into a fl agship species for climate change (see Ladle and Jepson 2010).Stork and Samways (1995) defined fl agship species as "popular charismatic species that serve as symbols and rallying points to stimulate conservation awareness and action."The important status of the golden toad is illustrated clearly by the following quotes from the websites of big NGOs and newspapers: conservation International's website stated "For Ticos, as Costa Rican natives are known, reducing greenhouse gas emissions and stabilizing the climate is personally important, as the extinction of their emblematic golden toad (Bufo [Incilius] periglenes) due to climate change and altered weather patterns is still fresh" (Conservation International 2011); Mr Andrew Kerr of the World Wildlife Fund stated "Global warming has already claimed its fi rst species, with Costa Rica's golden toad believed to have become extinct" (The Herald 2000); and The Independent (London), on September 18, 2006, stated "Human-induced climate change has already claimed its fi rst victims.The golden toad and the harlequin frog of Costa Rica have disappeared as a direct result of global warming" (The Independent 2006).
Conservation NGOs play an important role in determining conservation agendas, funding initiatives and interventions, liaising with politicians and policy makers, and, signifi cantly, communicating with the public (Jepson and Ladle 2010).However, information reported in scientifi c journals and further reported and possibly transformed in the public media can become asymmetrical, not necessarily tethered to scientifi cally recognised realities of uncertainty or improved analyses, and can take on a life of its own, even among relatively informed groups.More generally, our results provide insights into how reports in the peer-reviewed literature and the popular press can be closely related when peer-reviewed articles contain sensationalist or strident headlines and poorly related when not.The results also show how information from both sources changed over time and suggest ways excerpts of scientifi c information, once established in the public media, can be used further in important environmental, social, and political contexts around the world.
Nearly two decades ago Pechmann and Wilbur (1994) argued that there was insuffi cient information about amphibian populations to confirm the perceived worldwide decline of amphibian populations.More recently Salvidio (2009) re-analysed 16 amphibian populations monitored for more than 15 years, and reported that all of them have shown stable long-term population dynamics.This fi nding is intriguing, although we emphasise that 16 species and 15 years are insuffi cient sample sizes on a global scale for us to draw any clear conclusions.We see no reason to doubt that amphibians are a threatened taxonomic group as a whole, with many species in danger of extinction due to a range of environmental factors, including climate change (Collins and Storfer 2003;Wake and Vredenburg 2008).
When Pechmann and Wilbur (1994) raised the question of "Playing it safe or crying frog?", they were alluding to the potential negative repercussions of any incorrect reports of amphibian declines.Almost 20 years have passed since widespread amphibian declines were fi rst reported and population declines have been reported among many species, as they have for other taxonomic groups (e.g., Malcolm et al. 2006;Saino et al. 2011).Today, amphibians are no longer a forgotten group.Initiatives such as the Declining Amphibians Population Task Force (DAPTF) and others have increased general awareness and knowledge about amphibians substantially, although it remains the case that there is a lack of standardised monitoring data for most amphibian species/populations (e.g., Frías-Alvarez et al. 2010).
Communicating science via the popular media is diffi cult, yet critically important.As some authors have already emphasised, scientists, as sources in the media, have to think carefully about how the information they are giving is going to be interpreted (Friedman et al. 1999).The dynamics and goals of reporting information in the popular media often differ substantially from those of reporting research results in scientifi c journals (Schmidt 2009).This suggests that as a profession we need to strive harder yet to maintain the integrity of science in the public eye by stating clearly what we know and do not know when writing for a journal or communicating more directly with the public.Whilst scientists cannot control how the information they supply will be represented in the media and by interest groups and the polity, the following steps may reduce the probability of misrepresentation: careful drafting of press releases and responses to questions from the media; avoiding providing potentially misleading extrapolations or 'sound bites'; and providing succinct summaries of their articles written in layman's terms which are published on institutional websites as soon as media coverage is initiated.These summaries should include clear statements of the key assumptions and caveats of the research, the funding sources, and any confl icts of interest.We also recommend full use of digital communication platforms such as blogs (Ashlin and Ladle 2006) and social networking sites and that authors work closely with institutional press offi ces and take opportunities to participate in media training workshops and courses.

Figure 2
Figure 2 Proportional number of times per annum that each of the possible causes of amphibian decline were mentioned in news that included the golden toad.Not all the causes are necessarily related to the disappearance of the golden toad.
Fragmentation (346)There is a negative correlation between the abundance of frogs in the matrix and their vulnerability to fragmentation; however, results varied with fragment size and species traits Decrease in species abundance (Dixo and Metzger 2010) Pollution Pesticides and chemical wastes (499) Pesticides can cause variation in enzymatic levels among several tadpole species It appears that a conversion of native ecosystems to soybean crops may lead to increased ecological risks for anuran amphibians (Davidson 2004; Sparling and Fellers 2007; Rohr et al. 2008; Lajmanovich et al. 2010) Acid rain (24) Low pH has a negative effect on body size (Frisbie and Wyman 1995) Heavy metals (81): including iron, manganese, aluminum, mercury, cadmium, and bellyrium Cadmium inhibits acid secretion in stimulated frog gastric mucosa; Cu affects the liver metallothionein No apparent effect on population dynamics (Lefcort et al. 1998; Loumbourdis 2006; Cooper and Fortin 2010; Gerbino et al. 2010) Ultraviolet radiation (242) Lab experiments: UV-B has an overall negative effect in Ambystoma mexicanum.Proportion of deformed embryos varied through the year (Frías-Alvarez et al. 2010).Field experiments: "no differences in survival to hatching among UV-B-exposed and UV-B-shielded treatment" for Rana muscosa, Bufo canorus, or Pseudacris regilla (Vrenderburg et al. 2010) UV-B (280-315 nm) can be easily absorbed by living cells and causes DNA mutation and/ or cell death; induces lethal and sub-lethal effects on different ontogenetic phases of development 1.1% of the distribution of Ambystoma macrodactylum is exposed to lethal levels of radiation (Palen and Schindler 2010)Hatching success is unaffected by UV-B radiation.We suggest that UV-B radiation is an unlikely cause for declining populations of Rana aurora(Blaustein et al.  1996).Although UV-B radiation may not contribute to the population declines of all species, it may play a role in the population decline of some species (3 of 17) (Review of fi eld experiments;Blaustein et al. 1998) By itself, UV-B causes no signifi cant effects in Hyla chrysoselis and Rana blairi(Bruner et al. 2002)   Diseases Fungi (332): principally chytrids Focused principally in chytrids Mass mortality of several species in Central America(Lips 1998(Lips  , 1999;; Lips et al. 2005; but see Kilpatrik et al. 2010) Batrachochytrium dendrobatidis (Bd) destroys keratinised mouthparts in anuran tadpoles, which are essential for feeding Bd infection reduces foraging effi ciency of anuran tadpoles by altering feeding kinematics and reduces host fi tness (Venesky et al. 2010) Contd... [Downloaded free from http://www.conservationandsociety.org on Friday, February 14, 2014, IP: 192.38.113.119]|| Click here to download free Android application f this journal Appendix 2List of scientifi c articles including the terms "Bufo periglenes" and/or "golden toad" resulting from the searches in both Scopus and the ISI Web of Knowledge, as of April 2010.
Photochemistry and Photobiology

. Article Example of vulnerability towards chytridiomycosis Example of vulnerability towards climate change Example of climate change promoting chytridiomycosis No cause mentioned Example of APD but no related cause
[Downloaded

. Article Example of vulnerability towards chytridiomycosis Example of vulnerability towards climate change Example of climate change promoting chytridiomycosis No cause mentioned Example of APD but no related cause
[Downloaded

. Article Example of vulnerability towards chytridiomycosis Example of vulnerability towards climate change Example of climate change promoting chytridiomycosis No cause mentioned Example of APD but no related cause
[Downloaded