Modifying decisions depends on the cost of change, especially in sit-and-wait predators. Once situated, they should remain immobile, as position changes can disturb their traps or alert prey and predators. Antlion larvae (Myrmeleon spp.) dig conical pits in the soil to capture walking insects. The larva then ejects the prey carcass using strong, backward, pendular mandible movements. Therefore, their position on the pit influences the direction to expel carcasses and debris. We tested the hypothesis that antlion larvae orient at the bottom of their pits to avoid future movements, and that they decide their position based on environmental features such as the proximity of vertical obstacles and the ground’s slope. In the field, we dropped debris (rice grains) into pits. The expelled debris direction was consistent as larvae maintained their position at the bottom of their pit throughout time, and their position was independent of the pit's distance from a wall. In the lab, we induced larvae to build pits in narrow and sloped terraria. Larvae oriented themselves in positions that avoid expelling objects toward a wall or uphill, preventing them from bouncing off or rolling back into the pit. These results indicate that larvae use information about their surroundings to determine how to position themselves at the pit to avoid having to move if the pit gets compromised.

We conducted two studies: one field experiment to evaluate the distance and consistency of debris expulsion direction, and a combined field and laboratory experiment to examine larvae's orientation concerning environmental features.

I. In the field, in Yucantán, México, we experimentally determined the relationship between the expelling distance and the size of the larvae, and explored their consistency in the expelling direction. At the end of the dry season (May 2023), we randomly selected 45 different-sized antlion larvae pits. We measured their pit ratio with a caliper; an established proxy of larva size. Then, we carefully dropped half of a rice grain into each pit center from 10 cm above the pit, simulating the fall of debris and stimulating larvae to clean it.

To measure the consistency of the expelling direction, we repeated rice grain trials between 3 and 7 times in each pit. The number of repetitions depended on the pit status and larvae' behavior. The time between trials was approximately 5 minutes. We stopped the experiment when pits showed signs of landslides and disturbance, and when larvae did not respond for 90 s to our treatment. We then calculated a Consistency Index of Expelling Direction (CIED). We divided the area around the pit into 8 possible orientations of identical areas in function of the cardinal directions (i.e., N, NE, E, SE, S, SW, W, and NW). CIED was estimated as the number of the most frequent cardinal direction in which the rice was expelled, divided by the total number of repetitions. The CIED index takes a maximum value of 1 when all the expelling events are in the same cardinal direction (i.e., maximum consistency). We then divided the CIEDs into three categories of equal range (low, medium, and high consistency). 

II. In Sarapiquí, Costa Rica, we performed field and lab experiments, to evaluate whether the position of the larvae at the bottom of their pits was related to two environmental features (the proximity of walls and the slope of the ground).

First, we search for pits around buildings. We sorted pits as near a wall when the distance between the pit border and the wall was less than the pit ratio. At this distance, any object (debris or prey carcasses) expelled towards the wall direction might bounce off the walls and fall back into the pit. On the other hand, we sorted pits as far away from the wall when the distance between its border and the wall was larger than the pit diameter. At this distance, any object expelled out of the pit towards the wall is unlikely to bounce off the walls and fall back into the pit. We located 80 pits, measured their diameter, and their distance to the wall. Forty-six pits were categorized as near and 34 far away from the walls. After that, we carefully dropped half of a rice grain into each pit center as described above. The experiment simulated the fall of debris and stimulated cleaning by the antlion. Then, we characterized the expelling direction to the wall location using 4 subdivisions of equal area (toward the wall, opposite to the wall, right side, and left side of the wall). The orientation of the larva (i.e., where the larva's jaws were pointing) was considered the opposite of the expelling direction, given that larvae used their mandibles as catapults to throw out object out of their pits.

Second, in the lab, we evaluated whether larvae orient themselves to avoid throwing debris and/or carcasses toward nearby walls or uphill, preventing these objects from bouncing off or rolling back into their pits. We extracted 40 larvae by gently scooping the soil in the pit ~8 cm deep with a spoon. Larvae were randomly selected from those measured in the field experiment described above. We collected an equal number of larvae found near and far away from walls (n = 20 of each category). Subsequently, we transferred each larva to the laboratory and placed it at the center of an individual narrow container (rectangular boxes of 6 cm wide, 20 cm long, and 15 deep), filled with <2 mm fine-grain soil particles obtained from commercial sand separated with screen sieves. All containers had a 30-degree inclination. This design provided the larvae with only one position to avoid bouncing off or rolling back expelled objects: with its jaws oriented uphill. Objects expelled to the right and left sides of the container will likely bounce off and return to the pit due to the short distance between the pit and the container's border and those expelled uphill will probably roll back into the pit. Therefore, the best position to expel objects is if the larva positioned itself with its jaws uphill to throw out objects downhill. After 3 days, we measured pit diameter with a caliper and confirmed the larvae's position at the bottom of their pits. The larva's position was determined by dropping half of a rice grain into each pit center and determining its expelled direction in four equal-area options: right, left, uphill, and downhill. The position of the larvae was characterized as the opposite of the expelling direction, as previously described.