David Money and Jerald B. Johnson, Department of Biology
Our study takes advantage of a unique species pair within Brachyrhaphis, Brachyrhaphis roseni and B. terrabensis,, which provides insight into how divergence proceeds once speciation is complete. Brachyrhaphis roseni and B. terrabensis are sister species that occur in the same river drainages throughout northwestern Panama. A similar pattern occurs between these two species as occurs within other species of Brachyrhaphis, such that B. roseni lives in the presence of ambush predators (e.g., Hoplias microlepis and Gobiomorus dormitor) in lower elevation streams, and B. terrabensis, lives in largely predator-free, higher elevation streams. Therefore, using B. roseni and B. terrabensis as a model system, we are able to test whether patterns of divergence that have been documented within species occur in a similar way in between species comparisons, thus providing insight into the progression of behavioral divergence post-speciation.
The objectives of our study are as follows: (1) Evaluate divergence in boldness, activity, and exploration levels in sister species that occur in divergent predation environments; (2) Compare differences in divergence based on predation environment between different ontogenetic stages; and (3) Compare behavioral divergence among sexes. To accomplish these objectives, we used a behavioral assay that tested for several variables that can be used as indicators of boldness, activity, and exploration.
In order to test for differences between predation environments we conducted a two tier behavioral assay that examined the boldness, exploration and activity levels of each individual. We first conducted a test measuring the time to emerge into a novel environment, and then gave each individual eight minutes to explore the novel environment. The day prior to testing, we segregated an approximately equal number of juveniles, males and females of each species into separate 4-liter holding tanks in order to make individual members of each group were treated equally. For each trial, we carefully transferred a fish from its respective holding tank into the acclimation chamber, for a 2-minute acclimation period. We then lifted the trap door to allow passage into the novel tank. Once the fish entered into the experimental arena (i.e., once its entire body had crossed the threshold between the acclimation chamber and the novel arena), we lowered the trap door to prevent re-entrance into the acclimation chamber. We then allowed the fish to explore the novel environment for a total of eight minutes. After completion of their first trial, we placed each individual into a separate 2-liter holding tank, fed them, and then waited 24 hours before repeating the same assays. At the conclusion of the second trial, we removed the subject from the test arena, anesthetized it using MS-222, and took standard length (SL) and mass measurements.
From each trial we extracted four behavioral variables: (1) latency to emerge from the shelter into the novel environment; (2) proportion of time spent near shelter; (3) total proportion of time spent in motion; and (4) the rate of movement between the six zones. In total, we tested 17 juveniles, 21 males, and 20 females of B. roseni, and 19 juveniles, 19 males, and 18 females of B. terrabensis. All collecting and experiments were reviewed and approved by the Smithsonian Tropical Research Institute and Brigham Young University IACUC committees.
We found that juveniles of B. roseni and B. terrabensis did not differ in time to emerge or time spent in areas near shelter. However, juveniles of B. roseni spent a significantly higher proportion of time moving and had a significantly higher rate of movement than juveniles of B. terrabensis,. For adults, species did not differ in time to emerge. However, there was a significant sex effect and a significant species x sex interaction. This resulted from a dramatic increase in time to emerge in females of B. terrabensis relative to males. For proportion of time near shelter, we found a significant difference between species, with B. terrabensis spending more time in cells with shelter than B. roseni,. For proportion of time and rate of movement we found significant species and sex effects in adults, with a significant species x sex interaction for proportion of time moving. Overall, males spent more time moving and had a higher rate of movement than females, although this was primarily driven by a decrease in time moving and rate of movement in females of B. terrabensis.
In both juveniles and adults, we found clear species differences in behaviors between B. roseni (predation environment) and B. terrabensis (predator-free environment). Regarding boldness, although species did not significantly differ in latency to emerge from a shelter into a novel environment in juveniles or adults, B. roseni tended to emerge more quickly than B. terrabensis. These trends, although non-significant, appear to be consistent with previous work on poeciliids. In adults, B. roseni spent significantly more time out in exposed areas of the arena and away from possible shelters than did B. terrabensis. Since the most common predators with which B. roseni occurs are ambush predators (e.g., Hoplias microlepis), this behavior of spending more time in exposed areas, which is often considered to be a ‘risky’ behavior, might in fact be selected for in the presence of ambush predators that strike from less exposed areas. This difference was not present in juveniles, possibly suggesting that personalities might be playing a role here. Several personality studies have shown that predation environment correlates with personality. In other words, when predators are present, consistent prey behaviors are selected for, and when predators are not present, personalities do not evolve within the population. After inspecting mean times that individuals spent near a shelter, juveniles and adult B. roseni had very similar outcomes. However, juvenile B. terrabensis spent less time on average near shelter than adults. Therefore it might be possible that B. roseni does in fact have an evolved personality, and B. terrabensis does not. Although our experimental design differed from previous studies of Brachyrhaphis (e.g., use of shelter objects), the idea that B. roseni spends more time in exposed areas than B. terrabensis agrees with previous studies. Future work would benefit from evaluating the potential fitness consequences of these behaviors.
Brachyrhaphis roseni and B. terrabensis also differed in proportion of time spent active and the rate at which they explored the test arena. This pattern was consistent in both juveniles and adults. We found that adult Brachyrhaphis roseni were significantly more active and prone to explore than B. terrabensis, which is consistent with previous work with other poeciliids. Our work expanded upon previous work by evaluating these behaviors in juveniles. We found strong evidence that juveniles follow a similar pattern to that observed in adults, suggesting that these differences could persist across ontogeny. Differences in activity and exploration levels could be explained by differences in population density and resource availability in populations from different predation environments. Low population densities relative to B. terrabensis, and resulting differences in encounter rates with mates, have previously been elicited as possible explanations for increased activity levels in B. roseni. Furthermore, activity levels of populations occurring in different predation environments could be influenced by available resources, with resources typically being more abundant in low-elevation, high-predation streams relative to high-elevation, low-predation streams. In the case of adults, increased activity levels in high-predation environments could potentially result in an increased effort to find mates in habitats with lower population densities. However, this explanation would not apply to juvenile individuals. Therefore, it is possible that either different selective forces are acting on juveniles vs. adults (i.e., resource availability vs. mate availability, respectively), or that selection acting on either juveniles or adults favors stability of activity and exploration levels across ontogeny. Future work should evaluate ontogenetic stage-specific selection on behavioral traits to tease apart the underlying factors driving this pattern.
Our final objective of our study was to identify the sex differences in behavior, and especially how these sex differences interacted with species differences, based on predation environment. Brachyrhapis roseni was fairly consistent between sexes in all of our tests. These findings support a possible theory that predation environment might be a strong enough driver of selection that other involved factors such as sex roles might be undermined in evolution processes. On the other hand B. terrabensis behavior was much less consistent between sexes. Since they do not live in predator environments, these other processes of selection might have a much greater effect on evolved behavior. For example, male B. terrabensis were much more active than females. Without predators present, males might have a trade-off between being less active and conserving energy, and being more active and having more opportunities to mate. Therefore, they are less active than individuals from predation environments who must risk more in order to reproduce, but more active than females who will still have plenty of opportunities to mate with the more active males. In other words, these sex differences could be a result of differences in average personality within the population. While experiencing a high predation environment, B. roseni seems to be more consistent across ontogeny and between sexes. Conversely, while experiencing a low predation environment, B. terrabensis seems to be much less consistent across ontogeny and between sexes. Future work should focus on personality differences not only between individuals, but also on average personality differences between populations of divergent predation environments.
Our study supports other recent behavioral work by identifying several ecological and environmental processes that correlate with predation and tend to drive behavioral selection within a species. We found that populations experiencing predation environment tend to be more bold, active and explorative than populations that experience predation-free environments. We also observed these patterns within a species pair post speciation, which is consistent with other work done within species, and implicates behavior as a driver of speciation post derivation. We also observed several species x sex interactions in behavior, in which populations experiencing high predation environments were more consistent between sexes, while populations experiencing predation-free environments were more variable between sexes. In other words, predation divergence between species not only affected the superficial values of each expressed behavior, but also the deeper interactions within populations, greatly adding to the complexity of the system.