Long-term Memory of a Complex Foraging Task in Monitor Lizards (Reptilia: Squamata: Varanidae)
Procedural memory allows animals to solve previously encountered tasks over weeks, months, or years efficiently. Although thoroughly documented in vertebrate clades such as mammals and birds, studies of procedural memory in squamate reptiles are lacking. Filling the gap in knowledge regarding procedural memory in squamates is important to understanding the degree to which procedural memory is unique to birds and mammals, as it is related to their unique cognitive abilities. We tested for memory of a problem-solving task in two species of monitor lizard (Varanus spp.) and a beaded lizard (Heloderma sp.) after a 20-mo hiatus in exposure, representing approximately 25% of their ages at the time of testing. All the monitor lizards had lower initial latencies to solve the task upon re-exposure posthiatus than they had as naïve individuals during the prehiatus trials and reached minimum latencies in fewer trials than when previously tested. Our results indicate procedural memory of puzzle-solving behaviors on the time scales of years. Our results add to an emerging literature suggesting that squamate and other nonavian reptiles share a number of cognitive traits with birds and mammals, suggesting that such traits are far more widespread across taxa than previously recognized. We also discuss a framework for studying cognition in squamates that would allow tests of cognition across a great diversity of body forms and ecologies.ABSTRACT
The puzzle-feeder device. Mertens' Water Monitor (Varanus mertensi) investigating puzzle feeder device, in situ in its home enclosure. The dashed line marks the vertical plane of the device base. Once crossed, the latency time measures were initiated.
Latency to solve the foraging puzzle for each lizard. Each plot displays the individual's combined latency results from both the previous (Cooper et al. 2019) and present experiment with dotted vertical lines marking the 20-mo hiatus. Solid circles = solved trials, open circles = censored data where the individual did not solve the foraging puzzle within the 30-min trial period, and the solid lines = the Bayesian model fit with 95% credible intervals (shaded regions). The y intercepts estimate the initial latencies of each lizard's naïve attempt to solve the puzzle in 2017, and the slope is the decrease in latency as trials progressed. The model fit for H. charlesbogerti is not shown because it is outside the plotted area.
Contributor Notes
