Tuesday, November 17, 2015

Tails from the Cryptic Newt! Crowd-funding Campaign

One of my previous research assistants, Evan, is seeking your help! He tracked rattlesnakes with me back in 2014, and is now in the Master's program at the University of Texas, Rio Grande Valley. Evan is seeking donations to help fund his awesome research on black-spotted newts! Any and all donations help, even $5. Please share with others if you can.


Project Overview 

Black-spotted newts (Notophthalmus meridionalisare medium-sized salamanders within the Tamaulipan Biotic Province, ranging from San Antonio, Texas down to the southern end of Tamaulipas, Mexico. With a cryptic lifestyle, finding them is a difficult prospect. Typically, they are hiding underground in cracks or burrows estivating – possibly for years at a time – emerging at night during heavy rain events to migrate towards their ephemeral breeding ponds. These habits make finding black-spotted newts a time consuming and laborious effort.Black-spotted newts are listed as an endangered species by Texas Parks and Wildlife. Despite this, relatively little is known about them due in-part to their cryptic nature. What is known though, paints a tragic picture of a declining species. Fish introductions, cattle raising, and agricultural practices are a plight on their breeding sites. Click on the video below to learn more!

A potential newt breeding site

How Will the Study Be Conducted?

For the past 2 months, I have been locating possible newt breeding ponds throughout Texas and visiting nearby sites in southern Texas. Working in tandem with Jen Stabile, from the San Antonio Zoo, we will be visiting sites across Texas in the coming year. 
When newts are found, genetic material will be procured by taking a small (~1cm) clipping from the tip of the tail which will be stored in ethanol. Due to the cryptic nature of these newts, a new method called environmental DNA (eDNA) will be employed. This entails taking water samples from suspected newt locations and isolating genetic material from that sample. Genetic analysis will be performed using a recently designed technique called double digest RAD sequencing (ddRADseq). This method will generate a library of genetic sequences which will be compared between populations to identify population variation.
The study's budget

What Will My Contribution Be?

My project is focused on understanding the available variation within and between populations. When the data has been collected and analyzed, I will construct a distribution map depicting population divergence patterns. Sites with an abundance of newts will be given further consideration for future studies in migration habits to and from breeding ponds. With this information, future conservation efforts will have a greater chance of newt location. Further, genetic analysis will support if there is presence of a southern subspecies, as past studies suggest.

Go to project donation page to learn about awesome prizes!

Tuesday, October 6, 2015

Behavioral Thermoregulation in Sidewinder Rattlesnakes: Guest Post by Grace Freymiller

Grace is a new Master's student in the Clark Lab. She has been conducting research on desert rattlesnakes for the past three years, and she describes the work for her undergraduate thesis below:

I have been conducting research out in the Mojave Desert to understand how sidewinder rattlesnakes (Crotalus cerastes) behaviorally adjust their body temperature. Thermoregulation is the process of maintaining an individual’s internal body temperature within a specific range; all organisms have a range of body temperatures that they must stay within or else they risk dying. In humans, we thermoregulate via automatic responses to temperature change, such as shivering and sweating, which our bodies do naturally (ie. we have no control over these actions). Behavioral thermoregulation, on the other hand, encompasses active behaviors that are performed in order to regulate internal body temperature. Some examples of behavioral thermoregulation are huddling and moving from cool areas to warm areas (or vice versa).

A horned lizard behaviorally thermoregulating by basking in the sun.

Reptiles are ectotherms, meaning their body temperature is dependent on the temperature of the environment around them. Thus, they must behaviorally thermoregulate if they want to change their body temperature. For desert reptiles, thermoregulation is most often accomplished via movement across thermal gradients, such as moving from a cool burrow to a warm basking rock. Rattlesnakes, however, present an interesting scenario regarding behavioral thermoregulation: they are sit-and-wait predators. They will remain in ambush for hours at a time, during which they hardly move at all. Not only that, but they maintain their ability to strike both quickly and accurately throughout the duration of the night. 

Sidewinder (Crotalus cerastes), photo courtesy of Tim Garvey

One solution for sidewinders, which has not been well-documented, is through cratering. In 1992, Timothy Brown & Harvey Lillywhite coined the term “cratering” to describe the behavior that sidewinders often exhibit whereby they bury their outer coils in the sand when ambushing. The theory behind this behavior is that it keeps the snakes’ body temperatures warmer during the night and cooler in the morning because ground temperatures do not fluctuate as much as air temperatures do. One goal of my research was to answer the question “Does cratering provide thermal insulation for sidewinders?”

A cratered, ambushed sidewinder, photo courtesy of Tim Garvey

Two “craters” left behind in the sand after a snake left its ambush positions,
photo courtesy of Malachi Whitford

To answer my question, I surgically implanted sidewinders with temperature-sensitive transmitters, which send a signal to a hand-held receiver. The signal is just a continuous series of beeps that get louder as the person holding the receiver moves closer to the snake. Additionally, the time between each beep is dependent on the body temperature of the snake: if the snake is warmer, the time between the beeps is shorter. This conveniently functions as both a way to locate the snakes in the field, and as a way for me to determine internal body temperatures of the snakes.

I recorded the signal from these transmitters by placing a receiver in a cooler along with a voice recorder near an ambushed snake. I would record the snake’s body temperature for the whole night, providing me with a complete temperature profile for that snake on that night. I coupled these recordings with video footage so that I could link body temperature with behavior. I quantified the crater intensity of the snake (none/light, moderate, or heavy) using the video footage, then I examined how the snake’s body temperature related to both the air temperature and the ground temperature.

A temperature-sensitive radio transmitter 
(length-wise it’s about the size of a quarter) 

My temperature-recording set up. Inside the box is a receiver and a voice 
recorder, with the antenna on the outside.

If cratering is providing thermal insulation for the snakes, we expect that snakes with a heavier crater will have body temperatures more similar to the ground, and that snakes with a light crater will have body temperatures more similar to the air. However, my data does not support this. I found no relationship between the level of crater intensity and difference of body temperature to ambient temperatures.

Speckled rattlesnake (Crotalus mitchelli), photo courtesy of Steve Hein

So how do sidewinders do it? Previous research has demonstrated that this species of rattlesnake has a wide range of preferred body temperatures, which means that they can be fully functional at many different temperatures. One researcher found their preferred range to be 13.6 - 40.8°C, which is relatively large when compared to the sympatric speckled rattlesnake (Crotalus mitchelli), whose range was determined by the same researcher to be 18.8 – 39.3°C (Moore 1978). This means that as long as their body temperatures are within that range, sidewinders can allow their body temperatures to conform to ambient temperatures and do not need to behaviorally thermoregulate. 

One important thing to consider is that my research was conducted during the summer months, when even night temperatures are relatively warm, which could mask the effect of cratering during the night. During the summer, cratering would be most beneficial for snakes in the morning, when ambient temperatures begin to rise rapidly beginning at 5:30 AM. I am therefore continuing to explore cratering by determining whether snakes with a heavier crater stay out longer in the morning when compared to those with a light crater. Further research will try to determine why sidewinders crater if it is not a thermoregulatory behavior. One possibility is that it aids in camouflage, but additional research will need to be conducted to determine this.


Sunday, September 6, 2015

New Publication! How Do Young Squirrels Deal With Rattlesnakes?

The second chapter of my Ph.D. dissertation was recently published in the September issue of Behavioral Ecology and Sociobiology. Here's a little post summarizing my study in case you are not a journal subscriber ;-)

Only one more chapter to publish for me to graduate!!!

Besides being ridiculously cute, baby animals are prime 
targets for hungry predators

Life’s tough for a newborn animal.

Imagine you are a young ground squirrel pup. You are small, weak, inexperienced, and not fully developed, and you have just encountered a hungry rattlesnake.

This sounds frightening, but you, as a prey animal, have multiple opportunities to thwart predators during an encounter. For example, prey can detect predators quickly (before the predator detects them), and use behaviors to deter an attack or to escape an attack. Additionally, if prey are vulnerable at one of these stages, they might make up for it by having super-awesome defenses at another stage. For example, an animal might not be very good at detecting predators, but superb at escaping attacks. Thus, we need to examine how prey respond to predators across all stages of a predator encounter to understand where their vulnerabilities lie. 

The five main stages of a predatory encounter during which prey can use different defenses to avoid death. An encounter occurs 
when both parties are at a distance where they are both able to detect each other. An interaction occurs when the prey positively 
detects the predator and exhibits a behavioral response. Prey may try to deter an attack via signaling or active predator 
harassment. If a predator attacks and captures its prey, the prey can still escape after capture (not shown on this diagram). 
Flow chart adapted from Lima and Dill (1990).

Predators are a big deal for young animals because predators typically prefer to attack them over adults. Young can be born with almost completely functional antipredator defenses (especially those without parental care). However, more often than not, newborns learn from their mothers and neighbors which predator species are dangerous and how to handle them. This is especially true for social animals like squirrels.

The big question is: why don’t animals evolve to be born with functional defenses?? Needing experience is theoretically good for an animal because different predators might be dangerous at different times or in different areas. Thus, experience allows animals to modify their responses to the specific threats in their local area, but young are exceptionally vulnerable during the learning period. However, as I mentioned above, defenses used at one stage of a predator encounter can compensate for deficiencies at another stage, which can reduce the overall risk of predation on young.

The fees are definitely high for animals trying to learn how to
outwit predators. One mistake could mean death!
Few studies had examined age differences in antipredator defenses across multiple stages of a predator encounter, and so I set out to test this in the California ground squirrel-rattlesnake system. I examined whether squirrel pups differed from adults in antisnake behaviors during the detection, interaction, and attack stages of rattlesnake encounters. I specifically looked at squirrels’ ability to detect wild rattlesnakes, snake-directed behaviors after discovery of a snake, and responses to simulated rattlesnake strikes. I predicted that if I did not find age differences in a behavior, then the behavior is ‘un-learned’ and squirrels are born with a functional defense. If I did find age differences in a behavior, squirrel pups could have an inappropriate defense that requires experience to become fully functional, or squirrel pups could have an appropriate defense that is different from the adult form because it protects pups against specific risks only they experience.

I recorded wild squirrel-snake encounters in the natural habitat during two summer field seasons. I found that squirrel pups were not very good at finding rattlesnakes in their habitat. Adult squirrels detected snakes during 52% of all encounters, while pups only detected snakes in 21% of encounters. I found that during an interaction with a discovered snake, adult squirrels spent more time harassing the snake (by tail flagging, a behavior that deters the snake from striking and forces it to leave the area). In addition, if squirrels discovered a rattlesnake that was hidden in a refuge, adults were more likely than pups to investigate the snake’s refuge.

Example video of squirrel pups that are completely oblivious to a rattlesnake that's in 
ambush right next to their burrow. One of them is attacked by the snake (about 1 min in), 
but it uses its ninja skills to avoid envenomation:

I simulated rattlesnake strikes on squirrels and used high-speed video cameras to record their responses to these attacks. Squirrels, like many small mammals that are preyed upon by snakes, are secret ninjas that use aerial leaps to propel themselves outside the strike trajectory. We call this response an evasive leap. When attacked, squirrels can choose to either run away (called a scramble) or use an evasive leap. I found that pups were equally as likely as adults to use an evasive leap to escape a rattlesnake strike. However, pups had much slower reaction times to strikes than adults.  

Example video of the strike-simulating device and the two distinct flee modalities squirrels 
use to escape attacks (scramble vs. evasive leap): 

This study showed that squirrels’ ability to detect snakes improves with age. Squirrel pups should not wander around much and avoid areas with dense vegetation where rattlesnake set up ambush. Although pups are slow to react to rattlesnake strikes, they seem to use appropriate behaviors when dealing with a discovered snake. Squirrel pups do not approach snakes as closely as adults, and minimize time spent in close proximity to them. Thus, the behaviors young squirrels use at the interaction stage of a rattlesnake encounter appear to compensate for their deficiencies at the attack stage. On the other hand, pups are just plain bad at detecting snakes, an inappropriate defense that leaves them exceptionally vulnerable. Snake detection must be refined through learning over time.  

In my study, squirrel pups spent less time harassing rattlesnakes (fewer tail-flagging bouts) (panel a), were less likely to investigate
a snake's refuge (panel b), and had slower reaction times to surprise attacks (panel c). Taken from Putman et al. (2015).

Check out my recently published article here! 

Monday, April 20, 2015

Convergent Evolution in Small Mammal Antisnake Defenses

This post is part of a Reptile and Amphibian Blogging Network (RAmBlN) online event called #CrawliesConverge. We are writing on convergent evolution in reptiles and amphibians. Find our event schedule here. Follow on Twitter or Facebook.

Northern Pacific rattlesnake consuming a California ground squirrel pup.
Photo: B.J. Putman
Fitness is determined by two things: survival and the propagation of one’s DNA. Individuals that survive longer than others have more chances to reproduce and thus pass on their genes. Their success will determine the genetic makeup of the next generation (the foundation of natural selection). Predators act as one of the strongest selective forces in nature because once eaten, an individual is neither able to survive nor reproduce. It’s no surprise then that animals evolve many unique ways to deter predation.

Rattlesnakes and some other venomous snakes hunt in a stereotyped manner called ambush hunting. They remain stationary at a hunting site, patiently waiting for unsuspecting prey to wander by. They employ a rapid strike, during which they embed their fangs into the prey’s body. They inject venom into the prey, then typically release it and let it succumb to the venom before ingesting it. Because most venomous ambush-hunting snakes around the world do not drastically deviate from this general strategy, their prey experience almost the same predation pressures. Thus, prey* have independently figured out similar ways to avoid being eaten by ambush-hunting snakes.  

*by prey, I’m talking about small mammals (squirrels, gerbils, chipmunks, etc.). Snakes also consume other types of prey, but as far as I know, these prey do not exhibit the same antisnake defenses as small mammals. 

Only a few research groups are studying the predator-prey relationship between small mammals and snakes. The most well-known study systems involve the Cape ground squirrel and cobras/puff adders in South Africa, the kangaroo rat and rattlesnakes in the southwestern US, and the California ground squirrel and rattlesnakes in California (my system!).  However, we know many other small mammals around the world exhibit similar defenses from anecdotal reports. 

Left: a Cape ground squirrel harassing a cobra.  Right: the squirrel leaps away from a strike.
Photo: shockmansion.com 

Small mammals across the globe deter snake predation in three main ways:

1) Sending Signals

Many small mammals are not scared by snakes. They boldly approach them, investigate, and sometimes attack the snake! Typically, they will also spend a considerable amount of time repeatedly moving a body appendage in front of the snake. Most animals like squirrels and chipmunks wave their tails at the snake (Kobayashi 1987, Hersek and Owings 1993, Clark 2005), while others, like kangaroo rats and desert gerbils, drum their feet against the ground (Randall and Matocq 1997, Randall et al. 2000). Upon first glance, this seems like a pretty dumb thing to do in front of an animal that could kill you. However, these repetitive movements serve as warning signals to the snake, and actually deter it from attacking. Research from our lab has shown that squirrel tail-flagging tells the snake that it has been discovered (it has lost the element of surprise), and that the squirrel is prepared for an attack (i.e. if the snake strikes, it will likely miss) (Barbour and Clark 2012, Putman and Clark 2015). These signals could also inform other small mammals in the area of the snake’s presence, further degrading the value of its hunting site (because if everyone knows where the snake is, everyone is going to avoid the area). Some suggest that small mammals wave their tails to make themselves look like a larger more formidable opponent to the snake, but no study has yet tested this hypothesis.

Video of a kangaroo rat footdrumming at a sidewinder rattlesnake in the Mojave Desert.
Snake in burrow. Rat starts drumming 0:53 sec into video.
Video from the Clark Lab YouTube Channel.

Video of a Cape ground squirrel and mongoose harassing a cobra. Both exhibit aerial leaps when startled. 
Video from Smithsonian Channel. 

2) Ninja Reactions

When small mammals signal to snakes that they are ready and able to evade a strike, they are not lying. Thanks to snake predation, small mammals around the world have become ninja warriors, able to rapidly escape sticky situations using aerial acrobatics. This is a serious adaptation because a snake can strike at a velocity of 4.5 m/sec, meaning it could take a snake less than 70 milliseconds to strike a prey 30 cm away!  Small mammals need to use a response pathway that bypasses cerebral processing in order to react fast enough. Hence, it has been suggested that their responses to strikes are a type of startle response induced by acoustic stimuli rather than visual (meaning they respond to the sound of a snake strike and not the sight of one). The visual system responds via G-protein-coupled receptors, which are too slow to induce the speed of response we observe. The mechanoreceptors of the acoustic startle response are much faster and bypass cerebral processing. One study supports this claim by showing that only kangaroo rats with intact auditory systems were able to avoid rattlesnake strikes, while experimentally deafened rats could not (Webster 1962).

A quick response is important in avoiding a snake strike, but so is the type of escape you use. Successfully outrunning a snake strike is unlikely (see velocity above) – you would literally need to be the Flash to survive.  So instead of running away from strikes, small mammals have evolved the ability to leap vertically or horizontally away. This type of escape quickly propels the body of the small mammal away from the vector of a snake strike. Once snakes initiate an attack, their ability to alter their strike trajectory is limited, and so prey benefit more from displacing their bodies vertically or horizontally than by trying to outdistance the strike by moving within the same plane as the strike trajectory. Videos of these aerial leaps show that small mammals use their tails to contort themselves, often rotating their bodies near 180 degrees while midair. Evasive leaping is known to occur in squirrels, mongoose, and kangaroo rats. Kangaroo rats are the masters though, propelling themselves several body lengths upward when threatened with a strike.

Video showing the difference between squirrels running and leaping away from a 
simulated rattlesnake strike (spring-loaded cork). 
Video from Clark Lab YouTube Channel.

Video of a kangaroo rat leaping tremendously high in response to a strike.
Video from Clark Lab YouTube Channel.

3) Venom Resistance

Venom resistance levels of California ground squirrel 
blood paired with venom from different species of 
rattlesnake. Squirrels effectively inhibit the venom of
 C. oreganus and also  C. v. viridis, a close relative of 
C. oreganus (gray bars). Resistance against C. atrox 
venom is low (red bar) because California squirrels are 
not generally preyed upon by this snake species. 
Taken from Biardi et al. 2011.
Innate resistance to venom is what allows small mammals to closely approach and harass venomous snakes. Many small mammals are born with blood plasma factors that allow them to neutralize the effects of venom. Selection to minimize these effects is strong because even if an animal survives an attack, it must also minimize hemorrhage, tissue destruction, and disruptions to other bodily functions that follow envenomation because these could mess up the day-to-day activities of the animal. Thus, snakes and small mammals are often caught in an evolutionary arms race: snakes up their venom toxicity when the prey evolves a stronger defense, and this causes the adaptation cycle to repeat itself.
Several species of small mammal are known to harbor particularly high levels of venom resistance including the mongoose, opossum, woodrat, woodchuck, and some species of ground squirrel (see Ovadia and Kochva 1977, Perez et al. 1978, Poran andCoss 1990, Biardi and Coss 2011, Jansa and Voss 2011). Interestingly, the Cape ground squirrel which behaves similarly toward venomous snakes as the California ground squirrel is not resistant to either of its snake predators, the snouted cobra and the puff adder (Phillips et al. 2012). Findings like these help us understand the constraints associated with evolving such defenses.  

Many studies that have quantified venom resistance are flawed because they do not compare snakes and small mammals from the same area. For example, a study might examine how well squirrel blood inhibits venom by mixing it with venom pooled from snake species across the country, not just the species that only prey on the squirrels. Furthermore, if each population is in its own evolutionary arms race, even using venom of the appropriate snake species may provide us with an inaccurate measure of resistance – we have to use snakes and small mammals from the exact same population because prey evolve specific blood plasma properties to combat their local snakes, while snakes evolve specific venom properties to overcome the local prey’s resistance. Matt Holding at Ohio State University is doing exactly this with the California ground squirrel-Pacific
rattlesnake system, and others are on their way to improving our knowledge on this topic.


Snakes not only remove prey individuals from a population (rodent control), but impact their behaviors and physiology. The unique but similar defenses small mammals have evolved against snake predation demonstrate snakes’ importance as top level predators in ecosystems worldwide.  

The infographic below summarizes this blog post - distribute at will :) :) :)


Monday, March 30, 2015

In Search of the Coronado Island Rattlesnake

The famous arch of Cabo San Lucas
Baja California is my new obsession. The weather is perfect, the landscape is beautiful, and the ocean is always a stone’s throw away. I just got back from two trips past the border: (1) to Cabo for my honeymoon (where the screams of spring breakers soar through the air), and (2) to Southern Coronado Island in search of the elusive Coronado Island Rattlesnake.

Rattlesnakes live on several islands off the coast of Baja. Many of these snakes are the laid-back island counterparts of their mainland selves, and like other island inhabitants, many have drastically changed in body size from their mainland ancestors. For instance, Crotalus ruber lorenzoensis of San Lorenzo Island is a tiny descendent of the larger mainland Crotalus ruber. The Santa Catalina Island Rattlesnake (C. catalinensis) is one of the more famous species as it is the only known rattlesnake to “lose” its iconic rattle, although lorenzoensis and others seem to be on their way (Radcliffe and Maslin 1975). Poor thing only has one rattle segment to its name! Most of the Baja island rattlesnakes are endemic, which means they occur in one place and nowhere else (and this would be whichever island they inhabit). Although most of these snakes are not endangered, their endemic status is still a conservation concern because it means they have extremely restricted geographical ranges.  

The Catalina Island Rattlesnake (far left) is considered rattleless with only one small segment at the base of its tail. The two island
subspecies of ruber (middle and right) are on their way to becoming rattleless. Taken from Radcliffe and Maslin 1975.

The tiny island of Southern Coronado. Photo: B.J. Putman.

The Coronado Islands are about 15 miles South of San Diego and 8 miles from mainland Mexico. The largest of the four is Southern Coronado Island with just 1 square mile of land. The Coronado Island Rattlesnake (Crotalus oreganus caliginis) is the only snake species on the island*. Not much is known about this species and so my adviser at San Diego State University has collaborated with Dr. J. Jesus Sigala Rodriguez of Universidad Autonoma de Aguascalientes on a new project to learn more about its ecology. Last week, I went with him to the southernmost island to search for snakes that will become part of a long-term monitoring project. We drove from San Diego to Rosarito with minor hiccups other than being stopped by border patrol agents who could not comprehend the unusual amount of wooden cover boards in our trunk. From Rosarito, we took a panga boat to the island. In about 45 minutes we arrived to our destination. The island does not have a dock so our skilled boat driver slowly maneuvered the panga to the closest rocks, which we jumped onto from the boat.
Article correction: the night snake and gopher snake also live on Coronado Island

About to launch our boat in Rosarito, Mexico.

Although the island boasts 1 square mile of land, most of the land is inaccessible because it is so steep. We set off searching for snakes where we could and placed cover boards in areas that seemed promising. Alligator lizards (Elgaria multicarinata) were by far the most commonly encountered herp – we saw way too many to count! We also found a few skinks (Plestiodon skiltonianus). In all, our greatest find was a legless lizard (Anniella pulchra?).

We found 8 individuals of our target species, and not to brag, but this tiny girl found the most out of everyone in our group (brushes off shoulders). I found one of the rattlesnakes next to a dead alligator lizard. I assumed it was about to consume the lizard before I disturbed it. Interestingly though, the lizard’s eyes were missing which suggests it had been dead for some time (probably a day or two). We were thus unsure whether the snake had struck the lizard the day before and had just relocated it or whether the snake was scavenging it.

Left: The beautiful legless lizard.  Right: A rattlesnake found with an eyeless alligator lizard (sorry for my horrible photo editing skills)

Left: Processing a snake while enjoying the view.  Right: Drawing blood from a "tubed" individual.

We implanted a personal integrated transponder (PIT) tag into each snake we found. We also determined each snake’s sex, drew a blood sample, and recorded measurements on body size, rattle size, and mass. The Coronado Island Rattlesnake looks almost identical in pattern and coloration to the Southern Pacific Rattlesnake (C. oreganus helleri), its closest relative. In fact, the island snake is thought to have come from a mainland population of helleri. However, the island species dramatically differs from the mainland species in body size: it’s miniature in comparison. This may come as no surprise as island variants are often smaller than their mainland counterparts, a phenomenon known as Island Dwarfism.  

On top of a hill on the island. About to place down some cover boards.

Island dwarfism is an interesting phenomenon because it goes against Cope’s Rule, the prevailing trend in nature that organisms evolve toward larger body sizes. But islands provide special circumstances whereby they limit animals to a restricted area and also limit the amount of available resources. Hence, compared to giants, dwarfs are able to utilize limited resources more completely and are less likely to succumb to population crashes (Wassersug et al. 1979).

Left: A snake found coiled under a rock.  Right: Same snake tubed and ready for processing.

Island Gigantism also occurs, whereby animals grow in body size when isolated on islands. Gigantism is thought to result from many islands’ lack of large mammalian predators. Without these predators, animals are free to grow and exploit niches that were unavailable to them on the mainland. This is a phenomenon known as Predator Release.

Changes in available prey resources can also influence island dwellers’ body sizes. Across all island snake species, populations that are dwarfed tend to prey on lizards and populations that are giant tend to prey on colonies of nesting seabirds (Case 1978). For rattlesnakes that specialize on small mammal prey, islands lacking small mammals should lead to dwarfism and islands with prey equivalent to or larger than small mammals should lead to gigantism. Another interesting benefit larger-bodied viperid snakes have is their enhanced ability to fast compared to smaller snakes (Meik et al. 2010). Thus, gigantism benefits viperid snakes on islands where prey population sizes frequently fluctuate and/or where prey have relatively high extinction rates. 

Left: A specimen of the giant speckled rattlesnake of Angel de la Guarda.
Right: a specimen of the dwarf speckled rattlesnake of El Muerto Island.
Taken from Meik et al. 2010.
Almost all Baja California island rattlesnake populations are dwarfs. Only one, the Angel de la Guarda Island speckled rattlesnake (C. mitchelli angelensis), has undergone island gigantism. The Speckled Rattlesnake (Crotalus mitchelli) seems to love the island life boasting three descendants (C. m. angelensis, C. m. mitchelli, and C. m. muertensis) on 14 different islands. A study done by Meik et al. (2010) found that the body sizes of island populations of speckled rattlesnakes were best explained by island size (Meik et al. 2010). Smaller islands house smaller snakes and larger islands house larger snakes. The authors found that rattlesnakes typically tend toward dwarfism on islands that are smaller than about 20 square kilometers. In addition, rattlesnakes tended to dwarf on islands where the relative abundance of small lizards was greater than rodents. The authors then suggest that shifts to consuming larger prey (chuckwallas), fluctuating prey densities, and predator release likely resulted in gigantism for the speckled rattlesnakes on Angel de la Guarda Island.

Larger islands tend to house bigger snakes (SVL = snout-vent-length, a measure
of body size). Taken from Meik et al. 2010.

Dwarfism of the Coronado Island Rattlesnake is likely the result of small island size (less than 2 square kilometers!), and a greater reliance on lizard prey (pocket mice are the only known rodent inhabitant of the island). Research from the long-term monitoring program should provide quantitative data to support these theories. In all, islands are awesome (and beautiful) places that can act as “closed” ecosystems providing scientists the means to conduct unique experiments and find new discoveries.