INTRODUCTION

The Giant African Snail, Lissachatina fúlica (Bowdich, 1822), is one of the 100 world's worst invasive species (^{Lowe et al. 2004}, ^{Thiengo et al. 2007}). In the invaded areas, L. fúlica can cause impacts on health, agriculture, and the economy. In health, the snail can act as an intermediate host for nematode species of the genera Angiostrongylús and Aelúrostrongylús, which cause eosinophilic meningitis and abdominal angiostrongyliasis in humans, besides other diseases in domestic animals (^{Fischer et al. 2010}). L. fúlica is recognized as a generalist herbivore that feeds on various cultivated plant species (^{Thiengo et al. 2007}). Finally, the presence of L. fúlica entails an expense of the public herald for the potential consequences and control that must be carried out (^{Roda et al. 2018}). Therefore, ecological research on L. fúlica is necessary to mitigate the negative impacts of its presence through the construction of comprehensive management plans.

In the Neotropics, L. fúlica was first recorded in the 1980s in Martinique and Brazil (^{Mead and Palaci 1992}, ^{Santana-Teles et al. 1997}). By the beginning of the 21st century, it has already been recorded in Cuba (^{Vázquez and Sanchez 2014}), Venezuela (^{Martínez-Escarbassiere et al. 2008}), Colombia (^{De la Ossa-Lacayo et al. 2012}), Ecuador (^{Goldyn et al. 2017}) and Argentina (^{Gutierrez-Gregoric et al. 2011}). After recording L. fúlica, some countries have established regulations to control and manage the snail, besides inciting research to provide baseline knowledge (^{MAVDT 2011}). However, this control has been based on the manual collection and subsequent culling with chemical or physical means (^{Thiengo et al. 2007}), which requires ecological knowledge of local populations. As the species is widely distributed in the Neotropics (^{Darrigran et al. 2020}), a better understanding of the mollusk population dynamics in this region could help for enhancing control and management plans.

Density is one of the population parameters considered to assess whether an exotic population is established and the possible impact it produces. In L. fúlica, it is postulated that a population density higher than 10 ind/m² is already a cause for concern as an established population (^{De la Ossa et al. 2017}). This argument is because population density influences parameters related to fitness such as growth, fecundity, egg viability, and dispersal (^{Dickens et al. 2018}). Nevertheless, the relationship between population density and the impact of an invasive population is not linear (^{Jackson et al. 2015}) and can be affected by other ecological factors such as the trophic level (^{Bradley et al. 2019}). We would also expect that density is also affected by temperature and precipitation; however, as far as we know, these relationships have not been investigated for L. fúlica on a regional scale. In heterogeneous environments, such as those found in the Neotropics, understanding the relationships between density and climate should contribute to modeling the link between density and impact.

Here, we estimated the amplitude of climatic and anthropic variables where the population density of L. fúlica was recorded in the Neotropical region. Through the association of density values to climatic and anthropic variables, we aim to answer the following questions: How much does the population density of L. fúlica varies in the Neotropical region? Which are the main climatic variables affecting the density of L. fúlica in the Neotropical Region? How do these climatic variables, as well as the human footprint, affect snail density? This study is the first descriptive approach on the variation of the population density of L. fulica in the Neotropics, focusing on the intervals of climate and anthropic intervention where the species was recorded. We expect that this work will provide subsidies for decision-making for managing the giant African snail in the countries of the region.

MATERIALS AND METHODS

A directional search with the keyword "Achatina fúlica" was performed in Google Scholar in English, Spanish and Portuguese. This keyword corresponds to the old name for the species, as the genus change from Achatina to Lissachatina was recently accepted (https://www.marinespecies.org/aphia.php?p=taxdetails&id=88l469). We downloaded all documents presenting numerical values of population density in neotropical countries. Because some documents did not show the exact collection coordinates, the geographic coordinates of the localities recorded in each document were approximated in Google Earth, using the center of the locality recorded. To access the anthropic intervention in the Neotropics, we used the global human footprint index, which is expressed as a percentage representing the relative human influence in each terrestrial bi-ome (^{WCS 2005}). To access environmental parameters in the Neotropics, we used the nineteen current climate variables from WORLDCLIM 2.1 (https://www.worldclim.org/data/worldclim2l.html, resolution 2.5 arcminutes). We extracted the Human Footprint and climate values for locations with density records with the vegan (^{Oksanen et al. 2020}) and raster (^{Hijmans 2021}) packages in RStudio 4.1.0 (Supplementary material 1, Table S1).

Based on ^{Albuquerque et al. (2009)} and ^{Vogler et al. (2013)}, we identified Human Footprint, Annual Precipitation, Mean Temperature of the Coldest Quarter, and Temperature Seasonality as the most relevant predictors for L. fúlica establishment. Then, we categorized each of these variables as "low" and "high", taking as reference their median value in the collection sites, and constructed boxplots using density values of each point as the response variable.

To explore the population density variation of L. fúlica in the Neotropical region, we performed a principal component analysis (PCA). Variation was represented in a Cartesian space defined by the human footprint and the 19 climatic variables. Density values were classified as low (<0.001 to 3.35 ind/m²), medium (4.03 to 9.2 ind/m²), and high (10.45 to 150 ind/m²), taking into account the median of the data and the suggestion of 10 ind/m² in the study of ^{De la Ossa et al. (2017)}. The PCA and the graphical representation were elaborated with the packages FactoMineR (^{Lê et al. 2008}) and factoextra (^{Kassambara and Mundt 2020}) in RStudio 4.1.0.

RESULTS

In total, we found 22 papers estimating the population density of L. fúlica in the Neotropics. These papers record snail density in 36 localities covering six countries: Argentina, Brazil, Colombia, Ecuador, Venezuela, and Cuba, from 2004 to 2020 (Fig. 1a). For these records, the mean density was 11.55 ± 28.32 ind/m² and the median was 4.03 ind/m². The lowest and highest densities were reported in Cuba (Havana, 0.0002 ind/m²) and Venezuela (Andres Bello, 150 ind/m²). In most sites (80%) density values scored below 10 ind/m² (Table 1).

Figure 1 Density of Lissachatina fúlica in the Neotropical Region. a. Map showing the sites where density data were recorded. b. Density variation in sites under high and low values of Human Footprint, Annual Precipitation, Mean Temperature of the Coldest Quarter and Temperature Seasonality; the categorization of the environmental variables into high and low was based on their median. 

Table 1 Population density values at localities in the Neotropical region. 

Regarding the relevant predictors for L. fúlica establishment suggested by the literature, we found density records in the following ranges: Human Footprint from 21% to 93%, Annual Precipitation between 710 mm to 4438 mm, Mean Temperature of the Coldest Quarter between 13 °C to 27 °C, and Seasonality of Temperature between 3 °C to 40 °C. In general, we found no density pattern between low or high values of these variables (Fig. 1b).

The first two components of the PCA explained 63.5% of the variance of the environmental data (Supplementary material 2, Table S2). Mean Temperature of the Coldest Quarter (bio11) and Mean Temperature of the Driest Quarter (bioç)) had the largest contributions to the first component. In contrast, Annual Precipitation (bio12) and Precipitation of the Wettest Month (bio13) had the largest contributions to the second component. Human Footprint was the least contributing variable for both components (Supplementary material 2, Table S3). The Cartesian space defined by PC1 and PC2 shows that sites presenting low snail densities are distributed across the two components, indicating that different climatic combinations can maintain L. fúlica populations. On the other hand, most sites presenting medium snail densities show positive values of the first component, whereas most sites presenting high densities show negative values of the second component. This pattern suggests that density increase can be influenced by specific temperature and precipitation ranges (Fig. 2).

Figure 2 Principal Component Analysis (PCA) with the values of Human Footprint and climatic variables of the localities with recorded population density in the Neotropical region. Each point represents a locality, and its shape represents the recorded density (Triangle <0.001 to 3.35 ind/m², Square 4.03 to 9.2 ind/m², Circle 10.45 to 150 ind/m²) 

DISCUSSION

In this study, we found a high variation in the population density of L. fúlica in six Neotropical countries. Low densities of the species can be maintained in an array of environmental conditions, but density increase seems to be influenced by the Mean Temperature of the Coldest Quarter (bio11) and Annual Precipitation (bio12). The possible influence of bioll on L. fúlica density seems to be associated with establishment (^{Vogler et al. 2013}) and survival (Sharma and ^{Dickens 2018}). Annual Precipitation, in turn, may be related to physiological processes, like estivation (^{Rahman and Raut 2010}), that influence population dynamics. However, the low number of localities with density records does not capture all the environmental conditions that could explain the variation in density of the invasive mollusk.

Our data did not show the clear effects of human intervention on snail population density. The human footprint did not contribute significantly to the principal components, and the highest population density (150 ind/m²) was recorded in a site presenting the lowest value of human footprint (21%) (Supplementary material 1). In general, disturbed areas provide unique habitat opportunities and potential refugia for invasive alien species (^{Cadotte et al. 2017}) and may even promote adaptation (^{Borden and Flory 2021}). What is interesting in the case of L. fúlica is its ability to maintain populations in low percentages of the human footprint, which would turn conservation units and rural areas of countries into refuges for the species (^{Fischer et al. 2010}). Since L. fúlica can occur from low to high human intervention, we call attention to direct management strategies in rural human populations.

According to the latest distribution model of the species in South America (^{Vogler et al. 2013}), localities with low temperatures and high seasonal variation are less suitable for L. fúlica. However, in our study the highest density values (150 - 107.6 ind/m²) were found in localities with 4 °C and 34 °C temperature seasonality (Supplementary material 1). It is, therefore, possible to postulate that after establishment, L. fúlica can maintain high densities in high seasonality. In Andrés Bello, a tropical locality presenting a very weak seasonal climate, the high snail density could be explained by resource availability (^{Martínez-Escarbassiere et al. 2008}, ^{Herrera et al. 2016}). In the subtropical locality Puerto Iguazú, the high density could be explained by the sampling period: the collections were conducted in March, when temperatures around 31 °C and humidity around 70% (^{Gutiérrez-Gregoric et al. 2011}) represent optimal conditions for L. fúlica. This difference demonstrates that temporal monitoring is necessary to identify the influence of climate and resource availability on the viability of populations.

In conclusion, the population density of Lissachatina fúlica in the Neotropical region would be more influenced by climatic variables than by the degree of anthropogenic intervention. In this study, we postulate the Mean Temperature of the Coldest Quarter (bio11) and Annual Precipitation (bio12) as key climate variables influencing snail density, supporting the results found by ^{Albuquerque et al. (2009)} and ^{Vogler et al. (2013)} regarding L. fúlica establishment. We call attention to the ability of the species to maintain low population densities over wide ranges of environmental variables. Since at high population densities, there is a greater perception of damage (^{Jackson et al. 2015}), low-density localities are ignored when establishing control actions. In the future, a spatial and temporal expansion of local monitoring of this invasive species may provide sufficient tools to develop a relationship between density values and the impact produced by the species in the region.