Report Collective Motion and Cannibalism in Locust Migratory Bands

Department of Ecology and Evolutionary BiologyPrinceton UniversityPrinceton, New Jersey 08544SummaryPlagues of mass migrating insects such as locusts areestimated to affect the livelihood of one in ten people onthe planet [1]. Identification of generalities in the mecha-nisms underlying these mass movements will enhance ourunderstanding of animal migration and collective behaviorwhile potentially contributing to pest-management efforts.We provide evidence that coordinated mass migration injuvenile desertlocusts (Schistocerca gregaria) is influencedstrongly by cannibalistic interactions. Individuals in march-ing bands tend to bite others but risk being bitten them-selves. Reduction of individuals’ capacity to detect theapproachofothersfrombehindthroughabdominaldenerva-tion (1) decreases their probability to start moving, (2)dramaticallyreducesthemeanproportionofmovingindivid-uals in groups, and (3) significantly increases cannibalism.Similarly, occlusion of the rear visual field inhibits individ-uals’ propensity to march. Abdomen denervation did notinfluence the behavior of isolated locusts. When withingroups, abdominal biting and the sight of others approach-ing from behind triggers movement, creating an autocata-lytic feedback that results in directed mass migration. This‘‘forced march’’ driven by cannibalistic interactions sug-geststhatweneedtoreassessourviewofboththeselectionpressure and mechanism that can result in the coordinatedmotion of such large insect groups.Results and DiscussionEffect of Mechanosensory and Visual Stimulion Collective Motion in Schistocerca gregariaRecently, it has been demonstrated that alignment amongnearby moving neighbors allows locusts within aggregatesto align spontaneously and start marching once a thresholddensity of locusts is achieved [2]. This finding raises a keyquestion about the biological process that underlies suchcollective migration: Why should individuals align with neigh-bors? An important clue came from field studies of swarmingMormon crickets (Anabrus simplex) in the United States[3–5], where individuals’ motion is driven by the need to findnutrients such as protein and salt and where cannibalismwithin migratory bands is rife [3, 4]. If individuals fail to con-tinue moving they are likely to be attacked and risk becominganother cricket’s source of these essential resources. Impor-tantly, cannibalism in animals is a widespread and commonfeeding strategy [6], particularly so among grasshoppers,locusts,andMormoncrickets,inwhichitcanbeamajorcauseof mortality in the field [3, 7–9]. By conducting manipulativeexperiments combined with detailed analyses of individualand group behavior, we tested the prediction that cannibalismserves as a general mechanism for the transition from rela-tively disordered and benign aggregations to highly coordi-nated and mobile bands which are responsible for thedevastating impact of marching locusts [10].We experimentally manipulated the degree to which individ-ualscoulddetectmechanosensory(tactile)andvisualinforma-tion about conspecifics approaching from behind undercontrolled laboratory conditions. In addition, we investigatedthe importance of insects having sight of those insects ahead.Experiments were performed in a ring-shaped arena (80 cm indiameter) with a central dome, to restrict the perception ofother individuals on the opposite side of the arena (asdescribed in [2] and shown in Figure S1 available online). Indi-viduals in such arenas select collectively a (randomly deter-mined) rotational direction of travel. This simulates, in a rea-sonable space, the persistent collective marching of nymphsin sparse desert conditions [2]. Automated digital trackingsoftware [2] was used to obtain theposition, speed, and direc-tion of all individuals concurrently five times per s (see theExperimental Procedures and Figure S1).To manipulate the sense of physical contact from conspe-cificsapproachingfrombehind,weconstructedtwotreatmentcategories consisting of locusts that could sense tactile cuesfrom their abdomen and locusts that could not. To denervateindividuals, we severed the abdominal connectives of theventral nerve cord (immediately posterior to the metathoracicganglion, responsible for conveying sensory information fromabdominal exterceptors, such as cerci, to the brain and tho-racic centers [1]) through a small, hinged window cut in thecuticle (Figure S2). This denervates the majority of the abdo-men. All operated locusts were checked for healthy move-ment, and were left to recover for 20–30 min with food adlibitum. For control insects, we performed a sham operation,where the ventral connectives were surgically exposed butnot severed.Effect of Abdomen Denervation on the Behaviorof Isolated LocustsThe removal of sensation from the abdomen did not inherentlyinfluence individual motion. Tracking of the motion of bothcontrol and denervated individuals in isolation in the arenafor 4 hr revealed no difference in the proportion of time spent