Introduction

Human-induced disturbances cause global biodiversity loss, and one of the main drivers is biological invasions (Clavero and Garcia-Berthou, 2005). As human transport and commerce developed worldwide, various biota have migrated beyond their native distribution range following human activity (Hulme, 2009). Although most of the immigrants do not settle in the new range, some survive and breed excessively in the invaded area (Mack et al., 2000). These successful invaders can directly and indirectly interact with native species, and degrade native ecosystems (Lockwood et al., 2013). To conserve global biodiversity, it is essential to understand the population status of invasive species in each local area, especially unique ecosystems such as islands.

Island ecosystems have outstanding evolutionary distinctiveness, hence each insular biota is an essential element for global biodiversity (Banks and Dickman, 2007; Whittaker and Fernandez-Palacios, 2007). However, it is often the case that island ecosystems are especially vulnerable to invasive species because of the lack of co-evolutionary history with predators or toxic animals (Doherty et al., 2016). A previous study reported that island endemics accounted for 87% of extinct species (Doherty et al., 2016). The most impacted species are insular biota, indicating that understanding the population status of invasive species on islands should be a high priority.

Many previous studies have reported the presence or absence of invasive species on islands (Clavero and Garcia-Berthou, 2005; Lockwood et al., 2013). However, relatively few studies have evaluated population status after the invasion. The population that was introduced to an island can change dramatically because of environmental factors and species interactions (Kaji et al., 2009; Grant and Grant, 2006). Furthermore, it is common for multiple invasive species to be introduced to the same region, and they might have competitive interactions if they are closely related species or have a similar ecological niche. For example, introduced dingoes suppress introduced fox populations through competition for prey in Australia (Cupples et al., 2011), and introduced rats suppress introduced mouse populations through competition in New Zealand (Ruscoe et al., 2011). In such cases, quantifying the relative abundance of each invasive species would provide a significant insight into population dynamics and potential competitive interactions of the invasive species.

The cane toad (Rhinella marina) and Miyako toad (Bufo gargarizans miyakonis) were introduced to both Minami-daito Island and Kita-daito Island (Fig. 1) before World War II, in a harbor that has no native amphibians (Matsui, 1975; Ota and Toyama, 1992; Ota, 1999; Ota et al., 2004; Maenosono and Toda, 2007; Kita-daito village, 2017; Kita-daito village, 2019; Minami-daito village, 1990; Matsui and Maeda, 2018). Rhinella marina is a highly problematic invasive species that is native to southern America, and has a body length of 89–155 mm (Shine, 2010; Matsui and Maeda, 2018). This species was introduced to many areas worldwide to control sugar cane pests, and causes negative impacts on native predators via strong toxin defenses (Kidera et al., 2008; Shine, 2010). For example, native predators such as snakes, crocodiles and marsupials that attempt to eat R. marina are often poisoned due to a lack of co-evolutionary history of exposure to the toad's bufotoxins (Phillips and Shine, 2006; Doody et al., 2009; Kidera and Ota, 2008; Shine, 2010). Rhinella marina inhabits open lands close to human activity and sugar cane fields, and probably breeds all year around (intensively from December to January) in still water such as artificial ponds (Matsui and Maeda, 2018). Bufo gargarizans miyakonis is native to Miyako Island, Japan and has a body length of 61–119 mm (Matsui and Maeda, 2018). This species was also introduced to several islands to control sugar cane pests, and potentially affects native fauna (Ota and Toyama, 1992; Ota, 1999; Ota et al., 2004; Maenosono and Toda, 2007; Matsui and Maeda, 2018). Bufo gargarizans miyakonis inhabits sugar cane fields and breeds from September to March in still water (Matsui and Maeda, 2018). Both toad species feed on various terrestrial animals such as snails, centipedes, cockroaches, beetles, grasshoppers and ants (Matsui and Maeda, 2018). The breeding environment of both species is similar and the breeding period is long, hence they sometimes breed in the same place at the same time (Matsui and Maeda, 2018). Considering these ecological similarities, these species may have competitive interactions, however few studies have quantified their relative abundance. Quantifying the relative abundance of each species would clarify which species would be more abundant in the area, which might result from competition and/or environmental factors.

Fig. 1.

A map showing the study locations, Daito Islands (Minami-daito and Kita-daito Islands) approximately 360 km east from Okinawa Island, Japan.

Materials and Methods

Statistical analysis

We applied generalized linear models (GLM) to examine the relationship between the number of individuals in a line transect with species (R. marina vs B. g. miyakonis) and various environmental factors on each island. The error distribution was identified as a Gaussian distribution. We used the number of individuals in each line transect as a response variable and species, temperature, humidity, and wind speed as explanatory variables.

We also applied generalized linear models to examine the relationship between the number of individuals in each line transect with island (Minami-daito Island vs Kita-daito Island) and various environmental factors for each species. The error distribution was identified as a Gaussian distribution. We used the number of individuals for each species in each line transect as a response variable and island, temperature, humidity, and wind speed as explanatory variables. We obtained the data on temperature, humidity and wind speed from the free climate data provided by Japan Meteorological Agency ( https://www.jma.go.jp/jma/index.html last accessed on 15 Aug 2019). All statistical analyses were performed using R version 3.6.0 (R Development Core Team 2019).

Table 1.

Results of the number of individuals of Rhinella marina (R. m.) and Bufo gargarizans miyakonis (B. g. m.) and environmental factors on Minami-daito (MD) and Kita-daito (KD) islands.

Fig. 2.

Results of line transect surveys of (a) Rhinella marina (black points) and (b) Bufo gargarizans miyakonis (black triangles) on Minami-daito Island. Grey lines represent transect routes (Route 1–5) and thin lines represent water areas.

Results

The mean number of R. marina and B. g. miyakonis per 5 km line transect on Minami-daito Island was 121.2 and 3.0, respectively (Fig. 2a, b, Table 1). The mean number of R. marina and B. g. miyakonis per 5 km line transect on Kita-daito Island was 0 and 23.2, respectively (Fig. 3a, b, Table 1).

Fig. 3.

Results of line transect surveys of (a) Rhinella marina (black points) and (b) Bufo gargarizans miyakonis (black triangles) on Kita-daito Island. Grey lines represent transect routes (Route 1–6) and thin lines represent water areas.

The result of GLM analysis showed that the number of R. marina individuals per 5 km line transect was significantly greater than B. g. miyakonis on Minami-daito Island (coefficient estimate=118.2, p<0.001) (Fig. 4, Table 2a). In contrast, there was no significant difference in the number individuals between R. marina and B. g. miyakonis on Kita-daito Island (Fig. 4, Table 2b). Temperature, humidity, and wind speed had no significant effect on the number of toads on either island (Table 2a, b).

Table 2.

Results of generalized linear models examining the relationship between the number of individuals per 5 km line transect with species (Rhinella marina vs Bufo gargarizans miyakonis), temperature, humidity, and wind speed on (a) Minami-daito Island and (b) Kita-daito Island. Significance is represented in bold letters.

The GLM analysis showed that the number of R. marina individuals on Minami-daito Island was significantly greater than on Kita-daito Island (coefficient estimate=105.3, p<0.001) (Fig. 4, Table 3a). The analysis also showed that temperature and wind speed significantly increased the number of individuals observed (temperature: coefficient estimate=27.5, p<0.001; wind speed: coefficient estimate=15.9, p<0.01) (Table 3a). There was no significant difference in the number of B. g. miyakonis between Minami-daito and Kita-daito Islands (Fig. 4, Table 3b). Temperature, humidity, and wind speed had no significant effect on the number of B. g. miyakonis when comparing Minami-daito and Kita-daito Islands (Table 3b).

Fig. 4.

Box plots showing the number of individuals of Rhinella marina and Bufo gargarizans miyakonis that were observed along 5 km line transects on Minami-daito and Kita-daito Islands. The top bar represents the maximum value, and the bottom bar represents the minimum value excluding outliers. The upper and bottom bars of the box represent the upper and lower quartiles, respectively. The central bar of the box represents the median.

Table 3.

Results of generalized linear models examining the relationship between the number of individuals per 5 km line transect with island (Minami-daito Island vs Kita-daito Island), temperature, humidity, and wind speed of (a) Rhinella marina (b) Bufo gargarizans miyakonis. Significance is represented in bold letters.

Discussion

We found the relative abundance of R. marina was far greater than B. g. miyakonis on Minami-daito Island independent of environmental factors such as temperature and humidity. We also found the relative abundance of R. marina on Minami-daito Island was significantly greater than on Kita-daito Island independent of the environmental factors. In contrast, there was no significant difference in the relative abundance of B. g. miyakonis between Minami-daito Island and Kita daito-Island.

Our result of the field survey in 2019 on Minami-daito Island indicates that R. marina is likely to be more established than B. g. miyakonis on Minami-daito Island nowadays. However, interestingly, it was reported to be the opposite pattern 30 years ago: B. g. miyakonis was more abundant than R. marina on this island (Ota and Toyama, 1992). This temporal change of relative abundance of two invasive toads may suggest the instability of their populations on Minami-daito Island. Several studies have reported that population size, especially on islands, changes significantly over several decades. For example, the population of Sika deer (Cervus nippon) introduced to Nakanoshima Island, Hokkaido, Japan changed dramatically by degraded vegetation (Kaji et al., 2009), and the island population of Darwin's Finch (Geospiza fortis) changed drastically by drought and competition (Grant and Grant, 2006). Although there is no report of significant environmental change on Minami-daito Island and no study of species interaction between R. marina and B. g. miyakonis, there might be competitive interactions between them because their habitat, prey items, and breeding environment are all similar. Their potential competitive interactions might also influence the relative abundance of R. marina and B. g. miyakonis on Minami-daito Island. Moreover, poaching for the pet trade could be one of the potential factors influencing the change in relative abundance of toads on Minami-daito Island.

Although there was no statistically significant difference in the relative abundance between R. marina and B. g. miyakonis on Kita-daito Island, we found B. g. miyakonis easily, whereas we could not find any R. marina individuals on Kita-daito Island. This pattern is likely to be consistent with the previous study: B. g. miyakonis was more abundant than R. marina (Ota and Toyama, 1992).

Our results show that R. marina is abundant on Minami-daito Island whereas they are not abundant on Kita-daito Island. In contrast, B. g. miyakonis inhabit both Minami-daito Island and Kita-daito Island. Although the reason for the different pattern of relative abundance is not clear, environmental differences between these islands and ecological differences between these toads might be relevant. For example, the area of land mass and fresh water resource of Kita-daito Island is smaller than Minami daito-Island. Since the amount of water is different between these islands, water quality (e.g. water pollution by pesticide for sugar cane, salinity) might be different. Furthermore, these toad species may have a similar ecological niche and may use similar fresh water environment to breed, suggesting that they are potentially competing with one another as seen in other anurans (Komine et al., 2019). Bufo g. miyakonis is native to Miyako Island which is an uplifted atoll similar to the Daito Islands, hence B. g. miyakonis might have an evolved higher endurance against limited water resources on Kita-daito Island than R. marina. Revealing the potential competition and species differences of endurance against drought, water pollution and salinity would provide important insight to understand the patterns of abundance. Additionally, the relative abundance of invasive species may be related to their ecological impact (Watari et al., 2008; Komine et al., in press), hence further studies are needed to address the impact of these two invasive species on native ecosystems.

To enhance the validity of our conclusion, we should collect data in other seasons and years in the next study. Furthermore, we have no data on the number of individuals of each toad species initially introduced, or their initial introduction points. Initial population size might cause low genetic diversity, which might lead to population instability.