Differences in rheotactic responses contribute to divergent habitat use between parapatric lake and stream threespine stickleback

Abstract

Migration among populations is widely thought to undermine adaptive divergence, assuming gene flow arises from random movement of individuals. If individuals instead differ in dispersal behavior, phenotype-dependent dispersal can reduce the effective rate of gene flow or even facilitate divergence. For example, parapatric populations of lake and stream stickleback tend to actively avoid dispersing into the adjoining habitat. However, the behavioral basis of this nonrandom dispersal was previously unknown. Here, we show that lake and stream stickleback exhibit divergent rheotactic responses (behavioral response to currents). During the breeding season, wild-caught inlet stream stickleback were better than lake fish at maintaining position in currents, faced upstream more, and spent more time in low-current areas. As a result, stream fish expended significantly less energy in currents than did lake fish. These divergent rheotactic responses likely contribute to divergent habitat use by lake and stream stickleback. Although rheotactic differences were absent in nonbreeding fish, divergent behavior of breeding-season fish may suffice for assortative mating by breeding location. The resulting reproductive isolation between lake and stream fish may explain the fine-scale evolutionary differentiation in parapatric stickleback populations.

Sensory trait variation contributes to biased dispersal of threespine stickleback in flowing water

“…We measured rheotactic response and morphology of threespine stickleback ( Gasterosteus aculeatus , Linnaeus) from Blackwater Lake (northern Vancouver Island, BC, Canada) and its inlet stream. Previous studies demonstrated that stickleback from this lake and inlet stream exhibit nonrandom dispersal behaviour (Bolnick et al ., ) and divergent rheotactic responses (Jiang et al ., ).…”

“…The flow tank was equipped with two aquarium pumps generating unidirectional circular flow. Flow rates were within the natural range observed in the Blackwater inlet stream (Jiang et al ., ). The flow rate was the fastest in the outer part of the tank and the slowest in the inner part of the tank, allowing fish to chose among flow microenvironments.…”

“…In contrast, lake fish used high‐flow areas and faced downstream more often, resulting in greater net displacement downstream (‘negative rheotaxis’) despite repeated bursts of up‐current swimming. Note that this lake–stream difference in rheotaxis was only observed in wild fish caught during the breeding season, but not in laboratory‐reared or wild stickleback outside the breeding season (Jiang et al ., ). These lake–stream differences in rheotaxis could minimize migration between habitats (at least in the breeding season) and maintain population divergence (Bolnick et al ., ).…”

“…Examining the same lake–stream pair that showed biased dispersal, Jiang et al . () demonstrated that lake stickleback are more likely to move down‐current (negative rheotaxis) compared to stream fish. Specifically, stream fish more effectively used low‐flow boundary microenvironments and spent a large fraction of time facing up‐current, thereby remaining approximately stationary.…”

“…To test this hypothesis, we experimentally ablated stickleback neuromasts to determine whether they are necessary for a typical rheotactic response. Then, using both wild‐caught and laboratory‐reared lake and stream stickleback from a single lake–stream pair that was previously shown to exhibit biased dispersal and differences in rheotactic responses (Bolnick et al ., ; Jiang et al ., ), we tested for correlations between neuromast number and rheotactic responses. We also evaluated covariance between rheotaxis and body size, body shape, and pectoral fin morphology, which have previously been associated with stickleback swimming performance in stickleback (Walker, ; Walker & Westneat, ) and differ between lake and stream ecotypes.…”