In March 2014, the Asian toad Duttaphrynus melanostictus was reported from Madagascar's second largest city and main port Toamasina, raising immediate concerns about the invasive nature of the newly introduced toad and its environmental impact should it spread throughout the island. As part of a study on the feasibility of eradication, we conducted 516 interviews and 120 visual encounter surveys between April and November 2014. We found the toad to be widespread to the south and west of city center and estimate its minimum range to include an area of at least 108 km2. Social surveys indicate that the toad may have already been present for some years and potentially introduced prior to 2010, with the site of its introduction likely south of Toamasina near National Route 2 and the Ambatovy Plant. We discuss limitations of our survey methodology, proposed improvements for future work, and the implications of our results on eradication and control measures.
Introduction
Invasive species are among the greatest threats to biodiversity and of utmost concern to conservationists [12–3]. Within the Convention on Biological Diversity, the threat of alien invasive species is one of the 20 Aichi Biodiversity Targets for 2010–2020 [4]. Amphibians consistently rank among the most damaging vertebrate biotic invaders worldwide [5]. Amphibian species with the capacity to become invasive outside their native range are typically highly fecund, are easily transported, and have a generalist diet [6]. One such species is the Asian toad, Duttaphrynus melanostictus, a toad native to Asia recently detected in Madagascar, which is a significant pest in other tropical coastal areas [7–8].
In its native range, D. melanostictus inhabits a wide range of environments, including forests, agricultural land, and urban areas, where it is found up to 1,800-2,000 meters a.s.l. [9–10]. It reproduces throughout most of the year in tropical areas, with a peak during the monsoon season [11]. Populations in more temperate regions go through a period of inactivity during which reproduction stops when temperatures drop to 17°C [12]. In Cape Town, South Africa, an individual was found alive inside a shipping container transported from China [13], proving the adaptable nature of the species and how easily it can be moved unintentionally with imported goods. D. melanostictus also has a varied diet, with one study finding a range of invertebrates from more than 20 families within its stomach contents [14].
In March 2014, the scientific and conservation communities were alerted to the presence of the toad around Madagascar's main port and second largest city, Toamasina, causing widespread concern about the species' invasive nature and impact on native biodiversity. Questions were raised about the toad's origin, timing of the invasion, means of arrival, and the feasibility of control measures or eradication. Concerns especially focused on the potential effect that the toad could have on native fauna, an astounding 84% of which are endemic [15]. Only one other exotic amphibian species has been introduced to the island, the Indian tiger frog Hoplobatrachus tigerinus, and it is unclear when this species became established or the consequences of its introduction [16–17]. The newly introduced Asian toad on the other hand, presents a number of serious ecological impacts, including competition for resources, introduction of foreign pathogens, effects of predation, and the possibility that the toad's natural defensive toxins could harm naïve predators [18–19]. These concerns were heightened after the ecological niche of the species was modeled, revealing that Madagascar's climate, especially along the east coast, is ideal for the toad to spread [20].
Because D. melanostictus is poisonous, these concerns also extend to human health risks. D. melanostictus is implicated as the cause of poisoning in humans through consumption in Laos and has also been documented as the cause of death and cardiac arrest in children who have consumed toad tissue [21]. This health risk is elevated in the case of Madagascar, as its extensive rural populations to some extent rely on hunting and gathering for their sources of protein, their diet can include Anurans, and many people are unaware of the poisonous nature of the Asian toad. In addition to the Keomany [21] study there are a number of other cases of poisoning from consuming toads or their eggs [2223242526–27]. The health impacts of Asian toads in Madagascar may therefore be significant and lethal to humans, especially in communities where people eat frogs as a routine part of their diet.
When an invasive species is detected, determining its distribution is crucial in order to predict the rate at which a species will expand its range and to inform eradication or control strategies. These data, however, are often limited to randomly collected observations not recorded in a systematic manner, which affect the accuracy of such work [28]. Distributional data of invasive species, when possible, should therefore be collected methodically rather than through anecdotal reports. To this end, we describe our systematic surveys for D.melanostictus conducted from May to November 2014, as part of work to determine the feasibility of its eradication. We suggest a possible arrival time of the toad based on results from interviews with residents and offer a minimum distribution as of November 2014. We also discuss the limitations of our survey methodology, the importance of improving biosecurity measures in Toamasina, and the implications of the toad's distribution on potential control or eradication measures.
Methods
Two survey methods were used to assess the distribution and arrival time and location of the Asian toad in Madagascar: 1) a social investigation with local residents using Chef Fokontany (elected village officials) as the initial point of contact, and 2) visual encounter surveys at night.
Social Surveys
Initial interviews with residents took place on 30 April 2014 and 2 May 2014, with the help of 38 students from the Institut Supérieur des Sciences Environnemental et Développement Durable (ISSEDD). Students were divided into 17 groups of two to three people, and each group visited between two and 14 sites. At each site, the group first sought out the local Chef Fokontany and then asked permission before interviewing individual residents about1) if they had seen the toads at the site, 2) when they remembered first observing them, and 3) if they considered there to be many or few toads present. Informed consent was obtained from residents during interviews. Identification of D. melanostictus was not a concern because of recent media attention and the unique appearance of the species, which does not closely resemble any other amphibian in Madagascar (Fig. 1).
A second series of social surveys were conducted from 8–16 October 2014 by eight ISSEDD students divided into four teams. The methodology was the same as the earlier surveys, except that geographic coordinates were recorded using a Garmin GPSMap 60CSx GPS unit, the surveyors omitted question #3, and multiple residents were interviewed per site.
Visual Encounter Surveys
We carried out visual encounter surveys both to confirm reports from residents and to attempt to delineate the zone of incursion. Based on reports that the southern part of Toamasina surrounding the Ambatovy Plant already had a well-established population of toads, and considering the Indian Ocean to be the eastern-most limit, we plotted points expanding outwards to find locations on the edge where toads were absent or where densities were not yet high enough for the toad to easily be detected. These survey points were no less than 300 meters from each other and were selected for both their accessibility and their strategic location. Satellite (“BirdsEye”) imagery in Garmin BaseCamp was used to plot points, which were then loaded onto a Garmin eTrex20 and Garmin GPSMap 60CSx GPS unit.
Surveys were conducted by a team of four to six people (and on a few nights a group of ten). Between two and six points were surveyed in a single night, depending on their location and accessibility. The survey team began no earlier than 17:30 and continued no later than 21:00. Once a team arrived at the specified point, they walked together 300 paces in each cardinal direction (or as the landscape allowed) until a toad was observed. To save time, surveyors did not actively search for toads by turning over material or disturbing cover, but scanned the ground for active individuals using headlamps and flashlights. If a toad was observed, its presence was noted, and the team stopped searching and moved to a new point.
Results
516 residents were interviewed (141 in April-May, and 375 in October 2014) and 359 reported toads present at the site. The majority of residents interviewed reported first observing toads in 2014, while the earliest reports were from 2010 (Table 1). The five earliest reports were from Tanambao Verery (May and October 2010), Sosomita (2010), Andranokilaho (2011) and Tanandava (2011).
Table 1.
Residents who reported first observing toad per year given by number and percent. The last row “N/A” refers to interviews which did not result in an answer with a date as the time period (i.e. “a long time ago”).
From 120 visual encounter surveys conducted over 37 nights, we found toads at 48 sites (Appendix 1). They were found in all habitat types surveyed, from urban environments around houses and courtyards to more rural villages, agricultural land, and rice paddies, as well as mixed Eucalyptus forests, secondary vegetation, and neighboring grassy areas (Fig. 2). Toads qualitatively appeared to be most abundant around villages, and especially were seen around piles of garbage and debris.
Fig. 2.
The range of habitats where the Asian toad occurs around Toamasina. A) A waterway running through the city. B) Garbage pile and housing on outskirts of town. C) Grassy field and debris near center of distribution south along National Route 2. D) Temporary pond and surrounding scrubby vegetation around housing near industrial area south of the city. E) Aerial view showing a portion of the northeastern distribution of the toad. F) Rural land, Eucalyptus forest, and river in the southern portion of the known distribution.
The farthest point south that the toad was detected was Ambokarivo in Mahatsara (S18.28533; E49.32468) and the farthest north at a site approximately 500 meters north-northeast of Barikadimy (S18.13043; E49.37845). The toad was confirmed within less than 1 km of the ocean to the east, while the western-most points were Ambatavia (S18.19786; E49.29832) and Morafeno 2 (S18.15126; E49.30281), both located near the Ambatovy Tailings site. The toad was not detected near the center of the city, towards the airport, or farther north along National Route 5, or on National Route 2 south from Antanambao to Sandrangetana. Tracing the smallest possible convex polygon around the outermost sites of detection provides a minimum area of occupancy of 108 km2, with the centroid located along Route National 2 at S18.19206 E49.34256, approximately 1 km west of the Ambatovy Plant and the railway that runs south from the port through this area (Fig. 3 and Fig. 4).
Fig. 3.
Map of the distribution of the toad showing key features around Toamasina. A) Ambatovy tailings. B) Ambatovy plant. C) Ambatovy residential camp. D) city center. E) port. The starred flag corresponds to the centroid of the polygon and possible vicinity of introduction. The orange line is National Route 2 and National Route 5 and the yellow line is the railroad. The blue line is the Canal de Pangalanes. Map created using Mapbox ( www.mapbox.com
Fig. 4.
Map of all 120 visual encounter survey locations and the resulting distribution polygon for the Asian toad. Red flags correspond to sites where the toad was detected and green flags where the toad was not detected. Map created using Mapbox ( www.mapbox.com).
Discussion
Survey Considerations
It is important to note that interviews with the local population were done in a limited amount of time and on a voluntary basis, in the earliest case initiated before funding for surveys was secured. Voluntary and citizen-based biological research programs may have a greater possibility for bias and error than those carried out by researchers themselves [29–30]. While we made an effort to validate presence and absence data through visual encounter surveys following the first set of interviews, the estimated time of arrival could only be based on memory of individual residents. The majority of respondents reported first observing toads in 2014 and only a very few recalled seeing toads as early as 2010 or 2011. If the toad was introduced first prior to 2010, it would likely have occurred in low densities and at very few sites, and would therefore have been observed by very few people, if any. Only after 2–3 years of establishment and breeding would the toads become more detectable as their numbers grew and their range expanded.
It is interesting that there have been no reported observations of the toad in the city center or nearby at the port. The absence of toads at the port was substantiated by staff from Ambatovy who surveyed the area [31]. Toamasina is the largest port in Madagascar and therefore seems the likely entry location, but it does not seem to be an area where the toad has established a population despite suitable habitat being available. This suggests that the toads could have arrived at the port inside a shipping container and then been moved elsewhere inside the container to become established farther south, where all reports prior to 2014 are located. Indeed, the centroid of the distribution polygon is in an urban and industrial area south of Toamasina city, and assuming the toad has spread equally in all directions since its introduction, this could indicate where the toad initially established a population. The earliest social survey reports from 2010 and 2011 also are found around this center of distribution.
The sensitivity of our visual encounter survey methodology was likely low, and therefore the distribution that we provide is an absolute minimum and certainly underestimated. Furthermore, by the time this article is published the distribution of the toads will undoubtedly have expanded. Future visual encounter surveys could be improved by revisiting sites where toads were not detected over multiple nights, in varying weather conditions, as well as actively searching for individuals during the day by disturbing refuges and cover. It is important in such surveys to quantify detection probability to determine the accuracy of detection [32]. Future delimitation studies should also account for imperfect detection through methods such as modeled proportional occupancy [33]. Determining the edge of the toad's range is most important, especially for eradication programs or control measures, yet it is this area where toad densities are likely to be lowest and where they are most difficult to detect. For this reason, as is also recommended in the eradication feasibility report [34], it is vital that more sensitive survey methods be immediately tested and employed, such as investigating the use of tracking tunnels, acoustic monitoring, and environmental DNA.
While issues of security limited access to several private courtyards, most residents were helpful and granted access to survey within or through their property at night. Making contact with Chef Fokontany during the day prior to conducting surveys and carrying a copy of the permit authorizing our work helped facilitate entry to private property. Access was restricted, however, to several points. Notably, we were unable to conduct surveys on property of the Ambatovy mining operation, including the plant site, associated residential camps, and the mine tailings site, despite verbal communication suggesting the presence of the toad in these areas, which together encompass a substantial part of the known distribution of the toad. To remedy this, staff from Ambatovy conducted their own visual encounter surveys during June and July 2014 and confirmed reports on their property at a number of sites. While it is helpful to have information from a separate investigation, it will be necessary in the future to better coordinate surveys to ensure that all surveyors use a standard methodology and that data can be managed collectively. Furthermore, if an eradication effort is to be successful it will be essential to avoid splitting work between staff from the eradication program and private entities who may not follow the same protocol or have a background in invasive species removal.
Visual encounter surveys were also limited in that they only involved sites around Toamasina. We did not survey or search for toad populations outside the presumed site of introduction, though it seems possible with the amount of material that is transported out of Toamasina that satellite populations exist undetected elsewhere. Indeed, the toad was reported from the Salambona River in the quarters of Menagisa, Mahanoro more than 200 km south of Toamasina in May 2014, but thus far has not been confirmed, and the report seems somewhat dubious and without evidence [35]. Still, with the amount of trade emanating from Toamasina and the potential for toads to be unintentionally transported, there are likely to be undetected populations in the area outside of our mapped distribution. It is also important to point out that we detected the toad in several sites along the Canal de Pangalanes, a network of rivers and connected manmade waterways that extend more than 600 km southwards along much of the island's eastern coast. Assuming that the toad is able to disperse during its aquatic larval stage, it could potentially move unobstructed along the length of this canal rapidly.
Implications for conservation
Our results reveal that D. melanostictus is already widespread throughout and to the south and west of Toamasina and that its introduction to Madagascar possibly occurred four years or more prior to the government and scientific community being alerted to its presence. This has major implications for potential eradication or control measures, considering that for the former to be successful every individual toad must be destroyed at a rate faster than their ability to breed and recruit over an area now greater than 100 km2 [34]. The effort needed would be large, and the strategic and logistical challenges numerous. However, as pointed out in the recent eradication feasibility report [34] for which our distribution surveys were carried out, the consequences to the environment, human population and economy of doing nothing would also be severe and perhaps ethically unacceptable, even if cost and risk of failure are great.
Improving biosecurity at the port to prevent further introductions, both of toads and other potential invasive species, as well as to prevent toads from leaving Madagascar and spreading to other areas, is of utmost importance. An eradication effort cannot be attempted until biosecurity standards are improved [34]. Furthermore, in the likely event that eradication is not successful, it is crucial to improve biosecurity and prevent toads from spreading via exports from Madagascar to other areas of trade. In-country issues of biosecurity also threaten the viability and cost of eradication. Toads could easily be transported unintentionally around the island with goods and freight, and it seems likely they have already done so within their current distribution through the movement of garbage piles, in and around which toads are abundant.
Kull et al. [36] recently pointed out that invasive species have received far less attention in Madagascar than other threats facing the island's unique biodiversity. The introduction of D. melanostictus and failure to discover it until it was already widespread throughout Toamasina underscores this point, highlighting the need for conservation organizations and the scientific community to improve efforts to identify and address the threats presented by invasive alien species. With this in mind, we see an opportunity for a coordinated eradication effort to lead to critical methodological and strategic advances for the removal of pest amphibians globally, whilst resulting in improved policies and biosecurity procedures for the country of Madagascar.
Acknowledgements
This study was carried out with support from the Amphibian Survival Alliance. James Reardon kindly reviewed our manuscript before submission. Olivier Marquis helped edit our French abstract. We also acknowledge Christian Randrianantoandro, Angelica Crottini, Franco Andreone, Roger Daniel Randrianiaina, and Professor Eustache Miasa and his ISSEDD student volunteers for their assistance in organizing and contributing to efforts, as well as to staff from Ambatovy for sharing data from their own surveys. The Malagasy authorities of MEEMF/DVRN/DGF rapidly and considerately granted authorization for this work.
