Evolution in changing seas: The loss of plasticity under predator invasion and warming oceans

The impact of invasive predators during the early stages of invasion is often variable in space and time. Such variation is expected to initially favor plasticity in prey defenses but fixed defenses as invaders become established. Coincident with the range expansion of an invasive predatory crab in the Gulf of Maine we document rapid changes in shell thickness – a key defense against shell crushing predators – of an intertidal snail. Field experiments, conducted 20 years apart, revealed that temporal shifts in shell thickness were driven by the evolution of increased trait means and erosion of thickness plasticity. The virtual elimination of the trade-off in tissue mass that often accompanies thicker shells is consistent with the evolution of fixed defenses under increasingly certain predation risk.

Methods
Crab Surveys

Crab surveys were conducted at eight wave-protected sites to characterize regional variation of green crab abundance in the Gulf of Maine. Four sites were in the northern Gulf (Quoddy Region of Maine) while the remaining four were in the southern Gulf (Nahant to Cape Ann, Massachusetts). Surveys lasted for one hour, began and concluded within two hours of low tide, and were conducted monthly from July to October in 2017-2018 and 2021. Historical crab density data for 1973 and 2003 were obtained from previously published surveys.

Clinal Variation in L. obtusata Shell Thickness and Tissue Mass

Between 1995 and 1997, our previous work (29) collected samples of Littorina obtusata from 25 populations spanning GOM rocky intertidal habitats. Between 2017 and 2018, we collected samples from 22 of the same populations (snails were not present at 3 of the sites sampled in 1995-1997) to examine how phenotypic clines may have changed following increased ocean temperatures and increased green crab density in the northern GOM. For each population, snails (n = 50 per population) were haphazardly collected while attempting to maximize size range and returned to the laboratory for measurement of shell length and shell thickness with digital calipers. Mean shell thickness (hereafter, shell thickness) was calculated by taking the mean of whorl thickness and opposite whorl thickness. We then used a C-clamp to crack the shell of each snail and shell fragments were separated from soft tissue. Shell fragments and soft tissue were placed in separate aluminum trays and dried at 60°C for 48h before weighing on an analytical balance.

Reciprocal Transplant Experiment in the Field

To examine how shell thickness and its plasticity have changed after 20 years (1998-2018), in 2018 we repeated the reciprocal transplant experiment that was conducted in 1998. To allow a robust comparison of the two experiments, we carefully replicated all aspects of the 1998 experiment including using the same experimental chambers that were deployed into the field 20 years earlier. In early May 2018, we collected juvenile L. obtusata (5-6 mm in shell length) from a northern (Quoddy Head, Lubec, ME) and southern (Lobster Cove, Manchester, MA) site in the Gulf. All snails were individually tagged with a color-coded dot of permanent ink that was then sealed with cyanoacrylate glue. We measured initial shell length and shell thickness with digital calipers (± 0.01 mm) and shell mass and tissue mass (± 0.001 g) were estimated using a non-destructive weighing technique.

After completing initial measurements, we transported snails from both populations to the northern or southern site in mid-May. At each site, we placed six snails (hereafter, response snails) from a single population and approximately 60 g of brown algae (Ascophyllum nodosum) as food into 24 separate, replicate cylindrical containers (5-cm height x 10-cm diameter) that had mesh windows (mesh size = 3mm) to allow water flow. Hence, at each site 12 replicate containers housed snails from either the northern or southern population and 6 replicates for each population exposed snails to either the presence (Crab) or absence (No Crab) of predation risk. To create these risk treatments, each container stocked with response snails was secured beneath a similar container that was perforated on all sides and housed either (a) a mature male green crab (Crab) and 30 conspecific snails (hereafter, stimulus snails) or (b) just 30 stimulus snails (No Crab) to serve as a control. Each pair of stimulus-response containers was placed inside a large, replicate cylindrical chamber (11-cm height x 28-cm diameter) that had mesh windows (mesh size = 3mm) to permit water flow. These large chambers were anchored haphazardly in the mid-intertidal zone (~1.5 m MLW). Ocean temperature was monitored at 5-minute intervals during the experiment with dataloggers (Tidbits, model UTBI-001, Onset Computer Corp.) that were placed inside 3 replicate chambers at each site. Every 14 days we replaced stimulus snails in both the Crab and No Crab containers. For appropriate replicates, we also confirmed that crabs were alive; any dead crabs were replaced immediately. Overall, we had to replace 7 crabs at the northern site and 7 crabs at the southern site. At day 45, we replaced the Ascophyllum that served as food for response snails in all replicates. After 90 days in the field, all response snails were returned to the Northeastern University Marine Science Center (Nahant, Massachusetts) for measurement of final snail shell length, shell thickness, and tissue mass.