BEHAVIOR DURING SWARM MOVEMENT IN STELOPOLYBIA AREA TA (HYMENOPTERA: VESPIDAE)

When a colony of bees or wasps leaves an old nesting site and moves to a new one it is .taced with the problem of coordinating the movement of the individuals of the swarm so that all arrive at the new site. Scout honeybees (Apis mellifera) communicate the distance and direction of the new site by means of the same waggle .dance used to communica,te information about tood sources (yon Frisch, I967). Martin G. Naumann has observed swarm movement in several species of Neotropical social wasps and has described behavior that suggests that scent marks are used to guide swarm members to a new site (Naumann, I975). The purpose of this paper is to present the author’s own observations on behavior associated with swarm movement in Stelopolybia areata. The study was conducted during February 973 at the Estacidn de Biologia ’Los Tuxtlas’ near San Andres Tuxtla, Veracruz, Mexico.


INTRODUCTION
When a colony of bees or wasps leaves an old nesting site and moves to a new one it is .taced with the problem of coordinating the movement of the individuals of the swarm so that all arrive at the new site. Scout honeybees (Apis mellifera) communicate the distance and direction of the new site by means of the same waggle .dance used to communica,te information about tood sources (yon Frisch, I967). Martin G. Naumann has observed swarm movement in several species of Neotropical social wasps and has described behavior that suggests that scent marks are used to guide swarm members to a new site (Naumann, I975).
The purpose of this paper is to present the author's own observations on behavior associated with swarm movement in Stelopolybia areata. The study was conducted during February 973 at the Estacidn de Biologia 'Los Tuxtlas' near San Andres Tuxtla, Veracruz, Mexico.

BEHAVIOR ASSOCIATED WITH SWARM MOVEMENT
The colony in which the behavior was observed had been constructing a new nest two meters up in a bush along a small arroyo in a cow pasture. On the morning of February 4 the adult population of several thousand was found to have abandoned the nest and to have settled on a twig of a large tree about IO meters distant. The cause of the absconding was not certain, but several army ant (Eciton sp.) workers on the ground below the nest suggested that the nest may have been raided by a swarm of these ants earlier in the morning.
On the afternoon of the 4th and on each of the following three afternoons the swarm moved to a new site, providing an opportunity to observe its behavior. Data pertaining to each move are summarized in Table . The following description of the behavior of the wasps is extracted from the four days' observation.  In the mornings activity of the swarm was relatively low, though at least some wasps could always be seen flying within a meter or two of it. Observations of the swarm with a 6o X telescope during these hours indica.ted that much of the activity was due to the arrival and departure of foragers. The approach of a forager would cause several wasps on the swarm to reach out toward, the movement with their forelegs, while waving the .antennae. When a forager landed she would immediately told her wings, then often regurgitate liquid to these waiting workers. This behavior is identical to that observed on nests of the species. After such food exchanges the forager often disappeared into the midst of the swarm. The number of wasps engaged in foraging seemed to rise and fall with the appearance and disappearance of the sun in partly cloudy weather. Though data collected from other colonies at the same time of year suggest that 5-2% of the swarm population should have been queens (Jeanne,973), no queens were ever visible on the surface of the swarm. Their morphological distinctiveness from workers (Jeanne and Fagen, 974) shouldhave made them discernible had they been there. It is likely that queens remained near the center of the swarm.

1975]
Jeanne--8telopolybia :261 Toward midday workers could be seen flying about vegetation away rom the swarm. As the afternoon wore on, these became more and more evident, and more so in a particular direction away from the swarm. These wasps seemed attracted to prominent objects in the environment fenceposts, tall weeds or shrubs in the pasture, or leaves near the tips of twigs in tall second growth. In many instances they merely hovered a ew centimeters rom the object ("hovering"), but oten they landed and walked about slowly with the head down and gaster up, antennating the surface over which they moved ("landing").
During the course o an afternoon activity at the swarm varied.
Several times rapid increases in activity were observed during which great numbers of wasps took off and flew about in a large, diffuse cloud. During these times the numbers hovering at and landing on vegetation within o-2o meters of the swarm o.ten increased conspicuously. Within minutes, however, many o.]? these wasps returned to the swarm and the activity dropped to. its previous level.
As early as 43o (February 7), two hours and twenty minutes beCore the swarm began to move on that day, a new kind o.f behavior was observed. Some ot: the workers landing on the upper parts conspicuous objects walked rapidly over the surface, dragging the gaster continuously ("dragging"). The wings were usually buzzed, though sometimes this was interrupted with brief periods during which the wings were merely outstretched. The distance walked was usually in the range of o-5 cm, taking 2-3 seconds, though there was much variation. At the end of a run the worker immediately to.ok off.
Close observation o: these individuals indicated that either the apical margin o: the 5th (penultimate) sternite or the basal portion of the 6th was in contact with the substrate. Neither the tp o the gaster nor the sting made contact. The exposed surfaces o the and sixth sternites are more or less uniformly covered with short hairs, and don't appear markedly different from other sternites in this regard. The basal portion of the 6th sternite, normally overlapped by the .sth, is hairless and quite smooth. In workers this region is nticea.bly more convex than in queens. Though it was not possible to confrm durin observation, it may be this basal region that is in contact with the substrate. l..ater in the aiternoon the frequency ot: dragging increased, though not steadily. Often several workers could be seen performing it in a small area within a ew minutes, then none would be seen for 5 or 2o minutes By midafternoon the activities ot: hovering about, land-Psyche [June ing on, and dragging the gaster over objects were very obviously concentrated in one direction away from the swarm. This was invariably the direction the swarm would take when it moved. On one occasion I was able to trace these activities to 60 meters in the direction the swarm would take nearly an hour later.
Meanwhile, changes could be observed in behavior at the swarm. The level of general activity was greater over that of the morning hours. Foragers could still be seen coming and going and exchanging with wasps in the swarm, but a new kind of activity was also occurring with increasing requency. From time to. time a worker landed on the svarm or on the substrate at the edge o the swarm and ran agitatedly among the wasps. During these runs the wings were buzzed and the gaster was held low, possibly touching the substrate, though this could not be confirmed, and requently the gaster was waggled from side to side. Often the running wasp bumped into others, but without stopping. Sometimes, but not always, such behavior was followed by a noticeable increase in flight activity of the swarm as a whole.
Such periods of heightened activity, both at the swarm and away from it, occurred vith increasing frequency until the swarm finally began to move. This usually began quite suddenly as more and more wasps began to tgke off and fly in large arcs at increasing distances from the swarm. Within five minutes all but a few hundred xvasps had left the old site and were on their way toward the new one.
During movement o the swarm workers were requently seen dragging on prominent objects; this was performed by wasps moving both to and rom the new swarm site. Many wasps passing a given landmark, however, did not land, but merely hovered a ew centimeters downwind o it before flying on in the direction o the new swarm site. Others landed and walked over the surface o the object with their gas.ters raised.
The moving swarm was so diffuse that it would not catch the

DISCUSSION
The behavior (tecribed herein is very similar to what Naumann has observed in Polybia catillifex, P. oecodoma, 8telopolybia myrmecophila, Leipomeles dorsata, and Angiopolybia pallens (Naumann, 975). As Naumann concluded, it strongly suggests involvement o( a trail pheromone. The dragging behavior could well unction to deposit scent marks on prominent objects along the swarm route. Following swarm members may then pick up the scent by hovering downwind of such marks or by landing on marked objects and inspecting them with their antennae.

Psyche [June
What is needed nov is an experimental investigation of the whole matter, testing I) whether such a scent trail is indeed produced and followed, 2) how the swarm is finally stimulated to begin moving, and 3) how individuals recognize the new site when they arrive.