DESCRIPTIONS OF THE NESTS OF PSEUDOCHARTERGUS FUSCATUS AND STELOPOLYBIA TESTACEA , WITH A NOTE ON A PARASITE OF TESTACEA ( HYMENOPTERA , VESPIDAE ) . * BY

The nest built by a colony of wasps is the tangible product of complex behavior patterns which have evolved in response to a variety of factors. Since the nest is a. solid object which remains after the colony has gone, it provides a direct means of gaining insight into some of these factors. The remarkable architectural diversity of nests, especially among tropical species, suggests that the factors in question are many and varied. Beyond their importance in studies of the adaptations of individual species, nests must be taken into consideration in studies of the phylogeny of the social vespids. Richards and Richards (I95I) point out that the details of behavior as well as morphology must be carefully studied in deducing phylogenetic relations of each group so that those characters selected will combine to give the most consistent classification. Henri de Saussure (I853-8) was the first to. classify nests of social wasps according to their structure. Adolph Ducke, to. whom we are indebted for much of what we know about neotropical social wasps, attempted to construct a. phylogeny of the Vespidae based on Saussure’s classification (Ducke, I914), but his success was limited by inadequate taxonomic knowledge of the wasps themselves, and a lack of knowledge of the details of nest-building behavior. Ducke also made no attempt to understand the adaptive functions of the various features of nest structure. Though a great deal has been accomplished since Ducke’s time, especially through the work of Bequaert and of the Richardses, we still have much to. learn before we can hope to understand the evolution of nest structure in this family. The nests of roughly 30% of the, species of neotropical polistine wasps (exclusive of the genera Polistes and Mischocyttar;us) remain completely unknown, and many of the descriptions that do exist are cursory and note only general architecture. Almost nothing is known of the behavior associated with the detail.s of nest-building.

The nest built by a colony of wasps is the tangible product of complex behavior patterns which have evolved in response to a variety of factors.Since the nest is a. solid object which remains after the colony has gone, it provides a direct means of gaining insight into some of these factors.The remarkable architectural diversity of nests, especially among tropical species, suggests that the factors in question are many and varied.
Beyond their importance in studies of the adaptations of individual species, nests must be taken into consideration in studies of the phylogeny of the social vespids.Richards and Richards (I95I) point out that the details of behavior as well as morphology must be care- fully studied in deducing phylogenetic relations of each group so that those characters selected will combine to give the most consistent classification.Henri de Saussure (I853-8) was the first to.classify nests of social wasps according to their structure.Adolph Ducke, to.whom we are indebted for much of what we know about neotropical social wasps, attempted to construct a. phylogeny of the Vespidae based on Saussure's classification (Ducke,I914), but his success was limited by inadequate taxonomic knowledge of the wasps themselves, and a lack of knowledge of the details of nest-building behavior.
Ducke also made no attempt to understand the adaptive functions of the various features of nest structure.Though a great deal has been accomplished since Ducke's time, especially through the work of Bequaert and of the Richardses, we still have much to.learn before we can hope to understand the evolution of nest structure in this family.The nests of roughly 30% of the, species of neotropical polistine wasps (exclusive of the genera Polistes and Mischocyttar;us) remain completely unknown, and many of the descriptions that do exist are cursory and note only general architecture.Almost nothing is known of the behavior associated with the detail.s of nest-building.
*Manuscript received by the editor February 20, 1970.

Psyche [March
Pseudochartergus fuscatus has been recorded only rom the Amazon basin.Its nest is remarkable in that it is enclosed in a transparent envelope.This feature is shared with the congener P. chartergoides.P. fuscal,us also has cells that open upward, an extremely rare trait in the Vespidae.The nest of Stelopolybia testacea, a species that ranges from northern South America to Bolivia, is very similar in structure to those of 8. palli/oes and 8. cajennensis.The nest described here was unique, however, in having been constructed in a relatively exposed site, on the underside of a palm spathe.Moreover, it was one of the largest and most populous nest.,; of any social vespid ever taken in the American equatorial tropics.
A sample of the adult population of each nest has been deposited in the collection of the Museum of Comparative Zoology at Harvard University.The field numbers are 71 and 265 for the colonies of P. fuscatus and 8. testac,e'a, respectively.One specimen of the parasite of 8. testacea has also been placed in the MCZ collection.
The nest was located in an isolated tree growing in marshy soil at the upper end of an arm of a reservoir.The trunk of the tree would stand in shallow water during the rainy season (January through   June), when the reservoir is full.The tree was.about eight meters tall and twenty meters distant from tall second-growth forest, which surrounded it on three sides.The nest was built in the thick foliage at the end of a small branch about four meters from the ground.
The adult wasps were very timid, and the slight shaking of the nest as I collected it caused them all to retreat inside.Even more violent shaking after the branch was clipped would not bring them out to defend their brood.Only when the colony was placed inside the entrance.The parallel ridges in the broad sheet of transparent film to the right of the combs probably mark boundaries between successiv additions of the secretion.Slightly larger than natural size.distribution.The two, dark cells just above the five pupae near the center contain eggs of the third generation of brood.Large larvae are visible in surrounding cells.Some of the cells among the outer pupae are empty but still have fragments of cocoon attached to them; these are cells from which adults emerged after the nest was collected.1/2 times natural size.

Psyche
[March a plastic bag with cyanide did the adults fly rom the nest in a sudden burst.All o the 23o adults present were emales.Since the nest was collected during the day there were probably more out oraging.
The six leaves surrounding the nest were joined together with large amounts o a dry, transparent film, the resulting structure orming a watertight envelope around the nest, with a broad opening at the distal end (Figs. & 2).The film, probably a secretion produced by the adult wasps, resembled thick cellophane in that it was quite pliable, crackled when flexed and tore much like cello- phane.I wetted with water it softened but retained its orm.The ilm never covered lea suraces, but did bridge gaps o up to our centimeters between leaves.Each sheet o .theilm had a series of curved ridges o thicker material; these probably marked the bound- aries between successive applications o the material to the growing edge (Fig. 2).
The nest consisted of two.horizontal combs constructed of a fine- textured carton.The upper comb was the larger, being 4.5 cm wide by 6.5 cm long.It was suspended below a single leaf by three carton pillars averaging 5 mm in length.The largest of these was 3 mm thick, located near the center of the comb and attached to the leaf surace next to the midrib.The other two pillars, one on either side of the central one, were attached to the midrib of the lea.f.A t:ourth pillar joined one side of the comb to an adjacent leaf.The cells, of this comb opened downward.The second comb (2.0 cm wide by 4.o cm long) was directly below the upper one and fixed to the dorsal .surface of a smaller leaf by five pillars 5 mm long.The cells of this comb opened upward, though there were several irregularly-placed cells which opened to one side (Fig. I).The two combs were joined by a single pillar issuing from the face of the upper comb near its center and attaching directly to.a pillar at one end of the lower comb.A second pillar connected the face of the upper comb directly to the surface of the lower leaf next to the lower comb.The pillars were constructed entirely of carton.The cells of each comb averaged 3.o mm in diameter (measured from midpoints of opposite walls).Cells which had produced adults aver- aged 8.o mm in depth.The silken cocoons spun by the larvae were slightly domed and were not covered with bits of carton, as they are in at least some species of the genera Mischocyttarus, Clypearia and Epipona (Jeanne, unpublished data).
The upper comb (Fig. 3) had 288 cells in 25 rows.There were three generations of brood in this comb.Two .cells in the center of the comb contained eggs of the third generation, and next to these were ive second-generation pupae.Surrounding these were 58 cells containing meconia and second-generation larvae, decreasing in age away rom the center of the comb, and they in turn were surrounded by a ring of 90 cells containing meconia and second- generation eggs.Thirteen empty cells were scattered throughout these egg and larval cells.Around the second-generation egg cells were the remaining brood of the r.st generation-92 pupae sur- rounded by 25 larvae and three eggs.
The lower comb consisted of 52 cells.This comb was apparently younger than the upper one, and contained only two generations.The oldest part of this comb appeared to be at the proximal end (end nearest the lea petiole), where the end cell contained a meconium and an egg, probably of the second generation.Moving distally along the comb, there were, 37 pupae, three larvae and one egg of the qrst generation, with no.empty cells.
The entire nest, therefore, contained 97 eggs, 86 larvae, I34 pupae and 3 empty cells for a total o 330 .cells.Richards and Richards   (95) estimate the ages of nests in developmental units, one unit being the developmental period from egg to adult.By this measure the nest was just over two developmental units old.
There are three species of Pseudochartergus, and the nest o one, P. chartergoides, has previously been described by R. yon Ihering (9o4), A. Ducke (9o5) and J. Bequaert (938).It consists o a single, usually elongate comb, attached by pillars to the underside ot: a banana-like lea or a palm leaflet.The.comb. is enclosed by a second lea or leaflet t:astened under the (irst with a secretion similar to that described above for P. fuscatus.The nest o P. fuscatus differs rorn nests of P. chartergoides in having a second comb, and from nests of most other vespids in that the cells o this.comb open upward.To my knowledge the only recorded instance of a vespid nest with upward-opening cells is Ropalidia ltavopicta javanica rom Java (van der Vecht, I962).
The transparent (ilm enclosing the nest is likewise unusual but not unique to the genus.A. nest o Ropalidia opifex described by The term "generation of brood" as used here is defined such that the xth generation comprises all those brood in cells being used for the xth time.Since the adults of a colony may continue to add cells to the periphery of a comb that has developing brood in the center, several generations of brood may be present in a single comb at one time.number of pillars to the underside of a fern frond and enclosed in a "silk-like canopy resembling polythene in celour and degree of transparency" (p.70).The photographs of this nest reveal the striking resemblance of the canopy to the film described above for P. fuscatus, even down to the ridged thickenings.Van der Vecht mentions, however, that the covering was "very delicate" (p.7o), whereas that of P. fuscatus was quite tough.
To my knowledge Pseudochartergus and RoDalidia are the only genera to enclose their nests in a transparent film.However, wasps of the genera.Polistes, Mischocyttarus and probably others produce a similar substance which they apply with their mouthparts to build up the very tough petioles from which their nests are suspended (Rau, I933; Eberhard, I969).It is possible that these two sub- stances have a common glandular source.
Richards and Richards I95I classify the nest of Psmd:char;egus ckartergoides as stelocyttarous and gymnodomous; that is, the comb is attached by petioles to the substrate and is not surrounded by an envelope.But they observed no nests in the field and apparently relied on the description written by Ducke (I9Io), which makes no mention of a clear film joining the leaves surrounding the.nest.
Bequaert (I938), however, mentions the film, and it was consistently present in the nests of P. ckartergoides I found near Bel4m.Since the leaves and film form an envelope that surrounds the nest, much as do the more conventional carton envelopes of other genera, I am inclined to conclude that the nests of these two species can properly be called calyptodomous.

Stelopolybia (Gymnopolybia) testacea
This nest was collected on July aa, I968, in rain fo-et at Fazenda Taperinha, 4o km east of Santarm, Patti, B',zil  It was discovered where it had fallen onto a footpath.The nest had been suspended about ten meters above the ground beneath a large, woody spathe of an "inaja" palm (Maximiliana re:a Mart.).This inverted spoon-shaped structure, nearly a meter in length, had Fig. 4. The nest of Steloolybia testacea as it lay on the ground when discovered.The supporting spathe, to the left and rear, arches .over the combs, which lie stacked face up.The topmost comb, number 11, is folded back upon itself.There are numerous adult wasps on the spathe and combs.Fig. 5. 8. testacea nest reconstructed, as it would appear from beneath if the nest were in situ.The base of the spathe is at the right.The semi- cylindrical form of the nest is most pronounced in the older combs, at the top of the illustration.About 1/8 natural size.

Psyche
[March formed a natural roof over the nest, but the base of the spathe had so rotted that the weight of the nest caused it to break (Fig. 4)- The nest had lain on the ground for more than a week, but the adults had not abandoned it.The fall had torn the nest away from the spache, but the combs lay stacked on top of one another in ,the normal position relative to one another, so the nest could be roughly reconstructed (Fig. 5).
The nest was constructed of fine-textured carton of a uniform tan color.The cells averaged 5.9 mm in diameter (measured be- tween midpoints of opposite walls) and opened on the convex sides of the combs.The silken .caps of the pupal cocoons were slightly domed, and were not covered with bits of carton.The bottoms of the cells were convex, giving the backs of the combs a mammillated surface.
There were twelve combs, each hanging vertically from an edge by several pillars, and not surrounded by an envelope.The combs were constructed as concentric semi-cylinders around a vertical axis at one side of the spathe.Each comb was connected to a.dacent combs by many pillars (Fig. 6).The comb nearest the central axis was the smallest and apparently was the first to have been built.
Between combs 5 and 8 were three combs.Two of them, labelled 6 and 7, were at the proximal end of the nest (end nearest the base of the spathe; see Fig. 5).The third, called 7a, filled up most of the remaining space at the distal end of the nest.A vertical dis- continuity between the right and left sides of combs 4 and 5 (Fig. 8) suggests that these two combs were initially composed ot separate right and left portions which were fused as they were enlarged toward one another.A similar situation seemed to have occurred in the case of combs 6, 7 and 7a, except that 7a and 7 (or 6) were never used.The size and brood contents of each comb are given in Table I.
The presence of a meconium in a cell indicated that the .cellhad produced an adult.No attempt was made to determine the presence of more than one meconium in a used cell.Within a given comb the brood decreased in age trom the center outwards.Comb 8 Fig. 6.Combs 4-7 of the 8. lestacea nest showing how they are inter- connected by numerous pillars.About 1/2 natural size.Fig. 7. Comb of the 8. testacea nest in position on the spathe, combs 4-11 removed.The white caps of the pupal cocoons cover most of the central cells.
In Figs.7-9 the combs are shown in an upside down position; that is, the edge from which each comb is suspended from the spathe is at the bottom of the figure.The proximal end of each comb is to the right.

Psyche
[March showed this most clearly (Fig. 9).In comb 3 (Fig. 7), since the empty cells peripheral to the pupae had meconia, at least one com- plete generation of brood had been produced by this comb, and the pupae present belonged to a second generation.Since combs and 2 were probably older than 3, it can be assumed that comb had produced two complete generations and that the pupae of comb 2 were vestiges of the second generation.Comb 4 was then probably of approximately the same age as 3, and its pupae were of the second generation.Comb 5 had probably contained larvae of the second generation before the nest fell.Comb 7a was probably only slightly younger than 5, and had larvae of the second generation.Combs 6 and 7 had not produced any adults.Either they were constructed after 5, 7a and 8, or they were constructed in sequence and not u.sed.Since the cells peripheral to. the masses of capped cells in combs 8 and 9 (Fig. 9) were free of meconia, the pupae of these combs must have been of the first generation.Combs IO and I, free of meconia and with low cell walls, may have been empty or may have contained first-generation eggs and young larvae before the nest fell.
The entire nest contained a.n estimated 27,6oo cells, and of these approximately 3,ooo contained pupae.The nest was at least two- plus developmental units old, and possibly older.That is, it is possible that combs through 7a had produced more than two gen- erations o.f brood, and that comb 7a, may have produced at least one full generation before combs 8 through were even constructed.
It is interesting that nearly all the brood were in pupal stages, and that eggs and younger larvae were totally lacking.It may be that the fall of the nest caused the adults to remove the younger brood in order to.devote all their efforts to rearing the older brood to, maturity.Or it is possible that the fall of the nest had nothing to do with it, and that the colony was mature and was getting ready to swarm.In Panama, Rau (I933) collected an apparently mature colony o.f Stelopolybia pallipes vat.bequaerti, the brood of which Fig. 8. Combs 5, 6 and 7 of the S. test'acea nest in position on the spathe, with combs 7a-ll removed.The discontinuity in the rows of cells runs up the middle of comb 5. Combs 6 and 7 are nearly the same size and' cover the right side of comb 5. Fig. 9. Comb 8 of the S. testacea nest.The brood decreases in age from center outward.The empty cells in the center had all produced adults, and at the time the nest was collected adults were emerging from the innermost capped cells (note torn cell caps), indicating that these contained the oldest pupae.Newly-emerged adults can be seen resting on the comb.were almost all pupae.Rau concluded that the colony was about to swarm.
The. 8. testacea colony was collected at night, when all the adults were presumably on the nest, and very few escaped during: collection, so the 6,466 adults taken represent close to the total population of the colony at the time.However, due to the disruption caused when the nest fell, it is possible that the' population before the fall was significantly greater.
A sample of x,626 wasps was examined and found to comprise 5 males (7.%) and 5 females (92.9%).During the day ater the nest was collected one male and three females emerged rom combs 3 and 4, and one.male and 32 emales emerged from combs 8 and 9.This indicates that males are reared in both younger and older portions of the nest.During the four days following collection of the nest a total of 422 adults emerged from all combs, and o.f these 2o (4.7%) were males and 4o2 (95.3%) were females.
The fourteen species of 8telopolybia, all neotropical, fall into two subgenera, dngioolybia and Gymnopolybia.The nests of two of the three species of subgenus/lngiopolybia have been described (see Richards and Richards, 95, and the references they cite), and both are stelocyttarous, rectinidal and calyptodomous, meaning the combs are suspended horizontally one below the other by pillars and are surrounded by an envelope.
The nests of six of the eleven species of subgenus Gymnopolybia are known, though adequate descriptions seem to exist for only four.
All are stelocyttarous and gymnodomous, that is the combs are sus- pended by pillars a.nd are not surrounded by an envelope.They are built in hollow trees or other sheltered places.The architecture of the 8. (G.) testacea nest described here most closely resembles that of S. (G.) pallipes and 8. (G.) cajennensis, which build semi-cylin- drical interconnected combs hung vertically by their edges (Richards and Richards, 95r, and cited references). 8. (G.) vicina and 8. (G.) meridionalis hang their combs horizontally, each one sus- pended fro.m the one above by many paper pillars (R. von Ihering,  9o4).
The marked curving of the combs around a vertical axis seems to be the rule in 8. pallipes, cajennensis and probably testacea as well.The effect is to .stiffen the combs, making them much less likely to bend along a horizontal axis.Such a configuration may have been evolved to reduce nest damage due to wind when the nests are con- structed in exposed situations, as was the nest of 8. testacea.
As far as I am aware, the 6,466 adults of this colony make it the most populous colony of 8telopolybia ever collected, and it i,s among the largest colonies of any neotropicai wasp.The above-mentioned colony of 8. pallipes vat.bequaerti collected by Rau had 965 adults, and Richards and Richards (95) describe, taking a nest o.f 8. fulvo-fasciata in Guyana with 4o9 adults (many more escaped).
Eight .colonies of 8. pallens collected by Richards and Richards in  Guyana contained between 25 and 39o adults each.The Iargest colony of any New World polistine wa.sp on record is one of Brachyg.astramellifica collected near Brownsville, Texas, with about 5,ooo individuals (Schwarz, 929).Naumann (968) reported seeing a colony of the same.species with about IO,OOO adults (no locality was given).Richa.rds and Richards report taking a nest of Protopolybia pumila in Guyana with 7,o87 individuals.
These figures, of course, do not necessarily indicate maximum mature colony size.If the colony was collected during the day, many foragers may have been gone from the nest.In many cases in the literature a good portion of the adult population escaped during collection and were not counted.More fundamentally, the popu|a- Psyche [March tion size depends upon the age of the colony.For these reasons the figures cited can be used as only a very approximate guide to the mature size of a "typical" colony of each species.
Note on a parasite of Stelopolybia testacea On the second day following collection of the S. testacea nest, a parasite emerged from one of the .cappedcells of comb 8.It has been tentatively identified as Nomadina cisandina (Schulz) (Hymen- optera, Trigonalidae).It supert]cially resembled its host in form and color except that it was slightly smaller in size, had a relatively smaller head, and lacked the black markings on the gastral segments.The empty pupal skin of the S. testacea host remained between the cocoon of the parasite and the wall of the paper cell.The cap of the cell appeared normal and was evidently spun by the S. testacea larva.It is interesting that the adult parasite emerged from a section of comb rom which adult S. testacea were emerging at that time.2k second adult parasite of the same species was tound on the nest several days later.