Are tropical reptiles really declining? A six-year survey of snakes in Drake Bay, Costa Rica, and the role of temperature, rain and light

Introduction studies in the last two decades have found declining snake populations in both temperate and tropical sites, including informal reports from Drake Bay, Costa Rica. Objective to investigate if reports of decreasing snake populations in Drake Bay had a real basis, and if environmental factors, particularly temperature, rain and light, have played a role in that decrease. Methods we worked at Drake Bay from 2012 through 2017 and made over 4000 h of transect counts. Using head flashlights we surveyed a transect covered by lowland tropical rainforest at an altitude of 12–38 m above sea level, near the Agujas River, mostly at 1930–2200 hours. We counted all the snakes that we could see along the transect. Results snake counts increase from August to September and then decline rapidly. The May snakes/rainfall peaks coincide, but the second snake peak occurs one month before the rain peak; we counted more snakes in dry nights, with the exception of Imantodes cenchoa which was equally common despite rain conditions. We saw less Leptodeira septentrionalis on bright nights, but all other species were unaffected. Along the six years, the number of species with each diet type remained relatively constant, but the number of individuals declined sharply for those that feed on amphibians and reptiles. We report Rhadinella godmani, a highland species, at 12–38 m of altitude. Conclusion night field counts of snakes in Drake Bay, Costa Rica, show a strong decline from 2012 through 2017.


MATERIALS AND METHODS
We worked at Drake Bay, South Pacific of Costa Rica (N 08.69420-08.69490; W 083.67421-083.67495), from 2012 through 2017, and made over 4 000 h of transect counts of snakes. These counts were made while we were accompanied by small groups of tourists as part of our work as field guides in the area.
We hypothesized that climate changes may reduce local populations of some species and increase others; and that snakes that feed exclusively on amphibians or insect prey (which fluctuate strongly with rain or temperature), should be more affected than generalist species. Snake counts: Using head flashlights we surveyed a transect covered by lowland tropical rainforest, 12-38 m above sea level, near the Agujas River ( Fig. 1 in Digital Appendix 1). We counted all the snakes that we could see along the transect; identified them in situ (guides by Savage, 2002 andSolórzano, 2004), and photographed them for taxonomic corroboration (Digital Appendix 2). We worked mostly from 1930 to 2200 hours, but a few counts started at 1730 and ended at 2245 hours (sampling hours per date in Digital Appendix 2).
Precipitation rates and temperature: Precipitation rates and temperature data were kindly provided by the Instituto Meteorológico Nacional de Costa Rica from the nearest meteorological station at Rancho Quemado.
Note: Rancho Quemado, where the meteorological station is located, has an altitude of 240 m above sea level while the survey site in Drake Bay has an altitude of 12 to 38 m, so we expected a difference in temperature between the study site and the meteorological station. To assess this difference, we made temperature measurements directly in Drake Bay from February 2017 to August 2017 and compared them with Rancho Quemado for that same time period. We found that average temperatures in Drake Bay are 2 to 3 degrees Celsius higher than temperatures in Rancho Quemado, but that the trends along the timeline are the same. No precipitation records were available from January 2012 to July 2012, December 2016 to March 2017 and September 2017 to December 2017. No temperature records were available from January 2012 to July 2012 and from September 2017 to December 2017.
Moon and in situ rain: Instead of using published moon phase data, we recorded moonlight conditions on every trip directly in the trail because cloudy skies produce dark nights even when the moon is full. We used both official rain records and our own classification of rain condition during the trip (see Results).

Statistical analyses:
We analyzed the counts independently for each of the most common species, and pooled the rare species into an "others" category that we also analyzed ( Table 2 and Table 3 in Digital Appendix 1).
Ethical, conflict of interest and financial statements: the authors declare that they have fully complied with all pertinent ethical and legal requirements, both during the study and in the production of the manuscript; that there are no conflicts of interest of any kind; that all financial sources are fully and clearly stated in the acknowledgements section; and that they fully agree with the final edited version of the article. Tracie L. Stice and Gianfranco Gómez collected the data, Jose Pablo Barquero-González and Julián Monge-Nájera analyzed the data, all the authors participated in writing the manuscript. We did not need collecting permits because we did not collect any animals. A signed document with this statement has been filed in the journal archives.

Effect of diet:
Along the six years, the number of species with each diet type remained relatively constant, fluctuating by a couple of species each year ( Fig. 2 in Digital Appendix 1). However, the numbers of individuals changed visibly along the six-year study period depending on their main diet items. Species that mostly eat fish and invertebrates were always rare, and thus insufficient for us to see temporal trends; snakes that mostly feed on birds and mammals, which had larger populations, declined from 2012 through 2015. Finally, species that feed mostly on reptiles and amphibians were initially abundant but also have constantly declined over the sixyear period, despite some occasional population peaks ( Fig. 3 in Digital Appendix 1).
Annual patterns: When we compared counts of the five most frequent species with rainfall and temperature during the six-year study period, we noticed several trends (Fig. 4 in Digital Appendix 1). One is a weak increase in overall temperature along the study period. The other is that numerically dominant species have highly specific patterns but most started a decline since 2015. We observed a decrease in L. septentrionalis and I. cenchoa since 2015; E. sclateri is scarce (except for a peak from August to September of 2012 and 2014) but it has become even rarer since 2015; S. compresus was not seen after April 2016; finally, M. melanolomus had no clear tendency to grow or decline from 2012 to 2017, but showed a peak between August to September ( Fig. 4 in Digital Appendix 1).
Monthly pattern: Mean temperature does not change strongly along the year, but rainfall increases after February and reaches a maximum in October, while snake counts increase from August to September and then decline rapidly. The May snakes/rainfall peaks coincide, but the second snake peak occurs one month before the rain peak (Fig. 5 in Digital Appendix 1).

Ecological correlations:
The statistical significance values of ANOVA tests appear in Table 2 and Table 3 (Digital Appendix 1) and the counts/environmental condition relationships in the appendices, Effect of rain: We counted more snakes in dry nights, with the exception of I. cenchoa which was equally common despite rain conditions (Fig. 6 in Digital Appendix 1).

Effect of moonlight:
We saw less L. septentrionalis on bright nights, but all other species were unaffected so they are not presented in the figure (Fig. 7 in Digital Appendix 1).

Final considerations:
We did not see differences in leaf litter quantity from the beginning to the end of the study period (Digital Appendix 2). Our finding of R. godmani is unexpected because it is a highland species (see Discussion).

DISCUSSION
Food: Snake activity patterns depend on changes in food availability throughout the year (Henderson, Dixon, & Soini, 1978;Martins, 1994). Snake species that ambush their prey and rely on sit-and-wait strategies are particularly susceptible because of their low rate of food acquisition (Webb & Shine, 1998), just like specialist species, which are less likely to exploit alternative resources in response to shifting environmental conditions (Terborgh & Winter, 1980;Gaston, 1994). Counts of I. cenchoa, a snake that feeds mainly on reptiles, sharply fell in 2015 and 2017; E. sclateri, a specialist feeder on reptile eggs, declined in 2015 and almost disappeared in 2016-2017, possibly due to scarce prey. The fall in the numbers of L. septentrionalis matches the fall in the abundance of prey species like Boana rosenbergi, Smilisca phaeota and Agalychnis callidryas; nevertheless, other prey items for this species, like Rhinella horribilis and Craugastor fitizingeri, remained common (Gianfranco Gómez, personal observation).
If we consider that four of the five species affected by rainfall in our study have diets consisting primarily of amphibians, reptiles, or both, our results are consistent with those of a study in Mexico, where more rain lead to more amphibian activity and to increased populations of the snakes that feed on them (Duellman, 1958). Rhadinella godmani is a small leaf litter snake of uncertain diet and considered a highland species (Savage, 2002;Solórzano, 2004); we ignore if its presence at near sea level in Drake is in any way related to changes in environmental factors.
Temperature: Higher temperatures reduce the metabolic rate of snakes and restrain their activity and visibility (Seigel, Collins, & Novak, 1987;Zamora-Camacho, Moreno-Rueda, & Pleguezuelos, 2010;Rugiero, Milana, Petrozzi, Capula, & Luiselli, 2013). In Drake Bay, average monthly temperatures did not vary widely, so we are not surprised that there were no strong, generalized trends in species abundance that could be related with temperature. This matches previous work with tropical species (e.g. Shine & Madsen, 1996;Luiselli & Akani, 2002). Only one species, S. compressus, which needs fresh, shady habitats, was not recorded at all from our sampling site along 2016 and 2017.
Rain: There are reports of no correlation between counts of Neotropical snakes and rainfall (Henderson & Hoevers, 1977;Martins, 1994;Bernarde & Abe, 2006) and this is similar to our finding that the arboreal I. cenchoa was equally common despite rain conditions. Other species become more active with rain (Daltry, Ross, Thorpe, & Wüster, 1998;Oliveira & Martins, 2001;Morrison & Bolger, 2002), but we did not see any species more often in rainy nights in Drake, quite the opposite, Drake snake counts were higher in dry nights. Perhaps they avoid the cold rainwater (rain is cold in the tropics, too).
Moonlight: Snakes may increase activity in full moon nights, when prey are more visible (Lillywhite & Brischoux, 2012;Connoly & Orrock, 2018); for example, the tropical tree snake Boiga irregularis even moves to areas where moonlight is stronger (Campbell, Mackessy, & Clarke, 2008). However, Crotalus viridis avoids bright moonlight, possibly to escape detection by its predators (Clarke, Chopko, & Mackessy, 1996). In our study, the fewer sightings of L. septentrionalis on nights with strong moonlight may also mean that it is avoiding predation, but we saw all the other species at Drake with the same frequency in dark and illuminated nights.

Final remarks:
The clear decline in snake counts at Drake along the years might mean that their populations have decreased, that they moved out of sight or that they migrated to higher, cooler areas. They did not seem to avoid the transect because of our presence there (actually a few species remained constant in our counts along the years) and we do not think they found a suitable habitat at higher altitude because of human alteration of habitats around the reserve. We believe that the population decline in Drake is real and needs attention from the conservation authorities.