Dendrobatid poison frogs sequester lipophilic alkaloids from their arthropod prey to use as a form of chemical defense. Some dendrobatid frogs seasonally migrate between the leaf litter of the forest floor in the dry season to the canopy in the wet season, which may yield differences in prey (arthropods) and therefore alkaloid availability over space and time. Here, we document a seasonal vertical migration of Andinobates fulguritus (the yellow-bellied poison frog) from ground to canopy between dry and wet seasons. We observed turnover in alkaloid composition between seasons and found that dry season frogs contained a lower relative quantity of alkaloids; however, there was no change in alkaloid richness between seasons. The 77 alkaloids of 13 structural classes identified in this population appear to be derived mostly from mites and ants, though the two most common alkaloids were mite derived. Our observed shifts in defensive profiles are consistent with well-documented turnover in mite and ant communities between seasons and vertical strata. As climate change is expected to lengthen and strengthen dry seasons in many tropical regions, our results suggest that arboreal poison frogs forced to the ground for longer periods of time may see a shift in the abundance of alkaloids, possibly decreasing their defensive potential. This study provides further predictions for the wide-reaching effects of climate change, even as nuanced as charismatic poison frogs losing their poisons.

1. Study area

In central Panama, we surveyed an Isthmian-Atlantic Moist Forest located within the Esteban Alphonso Lee Natural Reserve (Lat 9.358555 : Lon 79.7029; 333 – 473 m a.s.l; Figure S1), which borders the Chagres and Portobelo National Parks.

2. Collection of A. fulguritus

We surveyed for A. fulguritus during the dry (mid-December – March) and wet seasons (April – mid-December) (Comita & Engelbrecht 2009, Basham & Scheffers 2020). Specifically, in the dry season frogs were captured between 4 February 2019 and 9 March 2019, and in the wet season between 2 July 2019 and 27 July 2019 (Figure S2). Thus, we attempted to standardize the amount of time experienced by individuals in both seasons before sampling, with individuals sampled between 2-3 months after the beginning of either season (Figure S2). Using survey methods in Scheffers et al. (2013) and Basham and Scheffers (2020), we conducted vertical, ground-to-canopy surveys for A. fulguritus, with each survey centered on a single Espavé tree, which is the principal habitat for A. fulguritus in this area (EB; unpublished data, Figure 1). A second researcher surveyed the ground in a 15 m radius around the tree for 1 hour searching through leaf litter, logs and other microhabitats (Heyer et al. 1994). During arboreal surveys, we searched for A. fulguritus in tree holes, moss, epiphytes and other microhabitat structures. For the purpose of this project we selected 13 adult individuals from each season collected from five sites (Table 1; Figure S1), for a total of 26 individual samples. Sites were located in contiguous forest at a range of distances from one another (closest sites, Tree IDs #114 to #111 = 19 m; furthest sites, Tree IDs #111 to #121 = 701 m, Figure S1). Samples were chosen to maximize sample size whilst representing sites equally across seasons, but due to the limitations of sampling and the need to avoid using the same individuals across seasons (checked by markings), there was a slight discrepancy in sampling across sites (Table 1). However, though tree #114 was represented by a single frog in the dry season, the trees #114 and #111 were closely located (19 m), thus reducing the potential for a significant difference in arthropod prey between selected sites.

Frogs were collected solely from the leaf litter beneath Espavé trees in the dry season and from the canopy of Espavé trees in the wet season. Espavé trees are host to a wide range of epiphytes, but were dominated by a small number of highly abundant moss, orchid, bromeliad, and ludovia spp epiphytes. Thus, the variation in arthropod prey availability across sites was reduced through the relative homogenization of vegetation structures and resources present across all sites.

3. Alkaloid Sampling

Captured frogs were weighed, measured (mean weight = 0.28 g; mean snout-vent length = 14.3 mm), brought to the research station, and allowed to rest overnight with ample moisture to maintain full hydration. Alkaloids were collected from each frog using a transcutaneous amphibian stimulator (TAS; Grant & Land, 2002), which applies a weak electric current to the skin, causing the secretion of the contents of their granular glands. Following the methods of Bolton et al. (2017), the TAS treatment (Frequency:  50 Hz; Pulse width: 2 ms; Amplitude: 9 V) was standardized among frogs. The frogs were then wiped with discs of absorbent bibulous paper to collect the secretions, which were deposited in 1 ml of 100 % methanol in glass vials with Teflon-lined caps. Frogs were fully recovered within 5 minutes, and were later released at the same point of capture. The TAS is a nonlethal method of collecting frog skin alkaloids, and previous studies have found no difference in the number and types of alkaloids collected using the TAS method when compared to the more traditional, yet lethal whole-skin extraction method (Clark et al. 2006, Hantak et al. 2013, Bolton et al. 2017, Schulte et al. 2017). However, based on studies of other small dendrobatids (Oophaga pumilio and Oophaga granulifera), the quantity of alkaloids extracted using the TAS are only proportional to the total quantity present in whole-skin extractions of alkaloids (Saporito, unpublished data). Therefore, the alkaloid quantities reported here (see Table S1) do not represent the total quantity of skin alkaloids present in A. fulguritus, but are instead proportional to the total quantity contained in each frog and should be compared based on their relative differences. We report quantity as micrograms of alkaloid per gram of frog.

4. Alkaloid extraction, identification, and quantification

For each sample, an internal standard of nicotine ((-)-nicotine 99%, Sigma-Aldrich, Milwaukee, Wisconsin) were added to 0.5 ml of the original MeOH extract. This extract was evaporated to dryness with N2, and then resuspended in 100 μl of 100 % methanol. Gas chromatography/mass spectrometry (GC/MS) was performed for each sample on a Varian Saturn 2100 T ion trap MS instrument coupled to a Varian 3900 GC with a 30 m x 0.25 mm i.d. Varian Factor Four VF-5 ms fused silica column. GC separation of alkaloids was achieved using a temperature program from 100 to 280 °C at a rate of 10 °C per minute with He as the carrier gas (1 ml/min). Each sample was analyzed in triplicate using electron impact MS, and once using chemical ionization MS with methanol as the CI reagent.

All alkaloids from each sample were identified using a combination of gas chromatography and mass spectrometry (GC-MS). Following the methods detailed in Hovey et al. (2018), individual alkaloids were quantified by comparison to the nicotine internal standard, and identified by comparing their retention times and mass spectral data with those of known alkaloids.

This data was collected by the Scheffers Lab at the University of Florida and the Saporito Lab at John Carrol University, OH, at significant cost of effort and resources.

If downloaded for use in publications we would appreciate a notification on its usage and would be glad to discuss potential research avenues and collaborations.

Please use this email for contact: brett.scheffers@ufl.edu

    
Metadata    

    
Alkaloids: The Alkaloids sheet contains all information pertaining to each specific alkloid that was recorded from the sample of 26 frogs

The information is broken down into the retention time, structural class, specific alkaloid name, the season from which it was sampled, the quanitity that was recorded in micro grams, a tentative label for arthropod origin, and the Individual Code for the relevant frog from which it was sampled

    
Frogs+Trees: The Frogs+Trees sheet contains information pertaining to the characteristics of each sampled frog (mass in grams), from when (Season and Date) and where (Height, and tree location as longitude and latitude) it was sampled. This data can be joined to the Alkaloids sheet using Individual_code as a linking variable