Filling the life history of the Hog-Islands Boa

Methods

Summary

We will conduct fieldwork on Cayos Cochinos Mayor and Menor from June to August 2019 and 2020 in order to find pregnant female Cayos Cochinos boas. June-August is the best period to find females close to parturition based on reports of females having been found gravid during June and are reported to give birth in late August and September on Cayos Cochinos Menor. Moreover, Boa imperator have been found to give birth from June to September on the islands off the coast of Belize (Boback and Carpenter, 2007). Once gravid females are found on the field, they will be held in individual field enclosures until parturition. Litter size and sex-ratio will be recorded from each female. Morphometrics (snout-vent length, tail length, head length, head width, and mass) will be taken from each neonate. The relationship between female SVL and litter mass among different populations of Boa imperator will be compared using an ANCOVA to provide an indication of the relative reproduction investment, assuming that energy budgets for boas from different populations are similar and only vary based on body size. In order to know if the Cayos Cochinos boas are using multipaternity as a reproduction strategy, we will collect DNA from blood samples (caudal vessels) from each pregnant female and from each neonates. Genomic DNA will be extracted and genotyped at eight microsatellites loci developed for Boa imperator (Bci-14, Bci-15, Bci-18, Bci-21, and Bci-23; see Booth et al., 2010; µsat-01, µsat-20 and µsat-36; see Tzika et al., 2007; 2009) known to be polymorphic in Cayos Cochinos boas (Greene, 2010). We will follow the DNA amplification protocol described by Stedman et al. (2016). We will use the program Gerud 2.0 to estimate the number of sires per litter (Jones, 2005).

The natural predators of the Cayos Cochinos boas remain unknown. To determine predation rate on snakes, scientists have used snake models made of soft and malleable materials (e.g. Webb and Whiting, 2006; Valkonen et al., 2011; Coombs, 2016). With the development of 3D-scanning and printing, models can be more biologically accurate and easy to build and manipulate (Bulté et al., 2017). In this study, we will use a 3D-scanner (Kinect camera, UTEP) to scan a preserved Cayos Cochinos boa specimen in order to create a digital model. This model will be printed (by Ultimaker, UTEP) in different sizes to represent different life stages of the snake (i.e. neonates, sub-adults, and adults). Models will be painted to be as biologically accurate as possible. Models will be placed in strategic places on both of the islands with a game camera monitoring potential predation events. These types of models can only be detected visually by predators. It is possible that some of the Cayos Cochinos boa predators use more chemical cues than visual ones. To be sure than we are attracting all potential predators, some of the 3D-printed models will be placed in aqueous extract of Boa imperator using the protocol described by Clark (2004). The predation rate of each model (with or without chemical cues) will be calculated and will provide a way to improve the use of models in predation studies. We will also paint some of the 3D-printed models following the pattern of the mainland Boa imperator morph to test the possible effect of the different color morph on predation pressure. As the 3D-printed snake models will be as realistic as possible, it is possible that humans will mistake them for real snakes and so, this method will also be used to assess poaching level on the islands.

Challenges

In order for this project to be successful we will need to find gravid females in the wild.  Gravid females have been encountered regularly during July-August on both islands, so finding these individuals during our study should not pose a problem. In order to determine if multiple-paternity exists within Cayos Cochinos boas, the micro-satellites markers being used need to present a high level of polymorphism (Jones, 2005). The eight micro-satellites we are planning to use are known to be polymorphic for Cayos Cochinos populations (Greene, 2010). In order to assess predation pressure on these snakes using 3D-printed snake models, the models have to be accurate enough to attract potential predators. Models are used by researchers to study prey-predators interactions (e.g. Webb and Whiting, 2006; Valkonen et al., 2011; Coombs, 2016) and can reach an efficiency of 50% (Coombs, 2016). We expect our models to be more efficient because they will be more realistic than those used historically (e.g. clay models; Coombs, 2016).