For insects that exhibit wing polyphenic development, abiotic and biotic signals dictate the adult wing morphology of the insect in an adaptive manner such that in stressful environments the formation of a flight-capable morph is favored and in low stress environments a flightless morph is favored. While there is a relatively large amount known about the environmental cues that dictate morph formation in wing polyphenic hemipterans like planthoppers and aphids, whether those cues dictate the same morphs in non-hemipteran (i.e. cricket) wing polyphenic species has not been explicitly investigated. To experimentally test the generality of environmental cue determination of wing polyphenism across taxa with diverse life histories, in this study we tested the importance of food quantity, parasitic infection, and tactile cues on wing morph determination in the wing polyphenic sand field cricket, Gryllus firmus. Our results also show that certain stress cues, such as severe diet quantity limitation and parasitic infection, actually led to an increase in the production of flightless morph. Based on these findings, our results suggest that physiological and genetic constraints are important to an organism’s ability to respond to environmental variation in an adaptive manner beyond simple life history trade-offs.

Cricket populations and rearing

We used G. firmus crickets from two laboratory populations originally collected from Gainesville, Florida. These colonies have been artificially selected for >40 generations by Dr. Anthony Zera to produce lines with increased production of either the flight-capable long-wing (LW) morph or the flight-incapable short-wing (SW) morph. Although these lines were selected for increased long-wing and short-wing morph formation, they still exhibit phenotypic plasticity for morph (i.e. the colony selected for long-wing can still produce short-winged adults). Additionally, we derived an unselected population from G. firmus collected from Tamarac, Florida in 2019. Because of the different origins of the individuals, we refer to these three separate lines using the following abbreviations: Selected Long-Wing (S–LW), Selected Short-Wing (S–SW), and Unselected (US). However, it is important to note that we were most interested in the question of the degree of influence that genetic and environmental factors have on morph determination at the individual level, rather than explicitly setting out to test differences between populations. Our goal was to identify individual level physiological responsiveness to the cues.  Populations were housed separately in large plastic bins in a climate-controlled room (26 – 28 ºC at 50 – 70% humidity) and provided ad libitum water and food. We used a standard lab diet of wheat germ, wheat bran, powdered milk and nutritional yeast (Zera and Larsen 2001).

Morphological ontogeny

To document morphological changes during development and to identify if there were any observable juvenile traits that indicate when the adult morph decision was made, crickets from the two selected populations (S–LW and S–SW) were separated within 24 hours of hatching. We used the morph-selected populations to ensure that roughly an equal number of long-wing and short-wing adults would be available to compare trait growth across the final juvenile instars to test for any differences between morphs. Each hatchling was assigned a random ID and housed individually in an 8 oz deli cup (Fabri-Kal® Pro-Kal™) containing an egg carton shelter and ad libitum food and water. We checked for nymphal molting events every 24 hours. To better visualize molting, we placed a small dot of non-toxic acrylic paint (Apple Barrel® 20591E Bright Magenta, 21476E Lime Tree) on the side of the abdomen; upon molting this paint was shed with the exoskeleton. After a nymph molted, we recorded the date and photographed the nymph from above using a Leica M125 automontage. This procedure was repeated for each instar until maturity was reached (8 – 11 nymphal instars). We waited 48 hours after adult eclosion to ensure full sclerotization of the exoskeleton before we euthanized individuals via freezing. Adults were photographed dorsally by a Nikon® D7500 mounted automontage both before and after forewing dissection. Upon completion of rearing, we randomly selected 25 long-wing and short-wing individuals of each sex to collect measurements of the terminal instar series and adult stages using ImageJ (Version 1.52s). For each of these 200 crickets, photographs from the adult, terminal juvenile instar (T), and the previous two juvenile instars (T-1 and T-2) were measured for head width, thorax width, abdomen length, total body length, forewing bud lengths, hindwing bud length, and ovipositor length in accordance with Supplemental Figure 1.

Food quantity treatments

To establish the quantity of food to provide crickets assigned to the high- and low- quantity food treatments, we measured food consumption of individuals in their T-1 juvenile instar. Thirty T-1 crickets of both sexes were removed from both the S–LW and S–SW lines within 24 hours of eclosing and placed in individual deli cups with 500 mg of food (enough to simulate ad libitum conditions), a cotton-plugged water vial, and a shelter. After 7 days, we removed the shelter, water, and frass to weigh the remaining uneaten food. This process was repeated weekly until adult eclosion. Average consumption per week was calculated as 203.5 mg (± 17.9 standard error).

To identify the impact of food quantity on morph determination, crickets were housed individually in deli cups within 24 hours of eclosion into the T-1 instar from each of the S–LW, S–SW, and US populations and assigned to the high-quantity or low-quantity food treatment. Crickets in the high-quantity treatment group were provided food ad libitum, whereas crickets assigned to the low-quantity group were provided only 58 mg of food every fourth day, which is approximately 50% of the average ad libitum consumption. All deli cups were provisioned with a shelter and ad libitum water. One day after adult eclosion, each cricket was weighed, euthanized by freezing, and wing morph recorded.

Density treatments

To identify the mechanism underlying the impact of local conspecific density on morph determination, crickets were housed individually in deli cups within 24 hours of eclosion into the T-1 instar from each of the S–LW, S–SW, and US populations and assigned to the simulated high- or low-density treatment. Nymphs assigned to the low-density treatment were housed in individual deli cups placed in a larger rearing container (Whitmor ClearVue® shoe box, 29.8 x 18.7 x 9.5cm) containing two egg carton shelters, ad libitum food, and two cotton-plugged water vials, but no crickets. Individuals assigned to the high-density treatment were housed in individual deli cups which possessed small airholes, placed in a larger rearing container containing two shelters, food, and two cotton-plugged water vials, five adult female short-wing crickets, and five adult male short-wing crickets. The crickets in the larger container simulating density were replenished on a weekly basis. We provided all focal crickets with ad libitum food, water, and shelter and monitored them daily for eclosion. One day after adult eclosion, each cricket was weighed, euthanized by freezing, and its wing morph recorded.

Parasitic infection treatments

Horsehair worm cysts (Paragordius varius) were obtained from a stream near Lincoln, Nebraska via collection of their intermediate aquatic snail host (Physa sp.). Snails were euthanized via freezing, then removed from their shells. Snail tissue was confirmed to be encysted prior to use in our experiment. To identify the impact of parasitic infection on morph determination, crickets were housed individually in deli cups within 24 hours of eclosion into the T-1 instar from each of the S–LW, S–SW, and US populations and assigned to the uninfected or infected treatment. Crickets were fasted for two days then fed either 5 mg of their standard diet or encysted snail tissue for the uninfected and infected treatments respectively. Crickets were monitored under red-light conditions for 15 minutes to verify food/snail consumption and then housed individually for the remainder of the study and monitored daily for eclosion. Upon molting, adult wing morph was recorded, crickets were euthanized via freezing and crickets in the infected treatment were dissected to confirm parasitic infection. Crickets that we did not detect a parasite in were excluded from the experiment as we could not distinguish between crickets that were never infected and those who successfully evaded parasite development via immune action. We did not collect weight data for this experiment given the inherent conflation between host and parasite weight.