Nesting Biology of the Solitary Wasp Pison argentatum
(Hymenoptera: Sphecidae) in Borneo and the Philippines
CHRISTOPHER K. STARR
Department of Life Sciences, University of the West Indies,
St Augustine, Trinidad & Tobago
ABSTRACT: Aspects of nesting biology of the widely distributed oriental solitary wasp Pison
argentatum Shuckard are reported from ﬁve localities in the Philippines and one in Borneo. Completed
nests comprised 1–9 cells, with average nest size differing between localities and nest substrates. Nests
based on hanging roots tended to be heavily plastered with mud pellets, while those on human-made
substrates were seldom plastered. Fully provisioned cells contained 7–21 spider prey. Over 75% of the
1003 prey recovered were Salticidae, while about 2% were Araneidae, the ﬁrst web-building spiders
recorded as P. argentatum prey. The remaining spiders were Lycosidae and Oxyopidae, like Salticidae
mostly long-sighted non-web-building hunters, but also new prey family records for this wasp. P.
argentatum offspring from three samples showed sex ratios (male/female) of 0.41, 1.53 and 2.40.
KEY WORDS: Nest structure, prey, Pison, Sphecidae
Two genera of trypoxylonine sphecid wasps are found in East Asia. Both Trypoxylon and
Pison use mud in making their nests, in which they store spiders as brood food (Bohart and
Menke, 1976; Iwata, 1976). Pison is a cosmopolitan genus of about 200 known species
(Menke, 1988). Of these, P. argentatum Shuckard has the broadest range, being found from
Hawaii to Madagascar and widespread through South Asia (Baltazar, 1966; Bohart and
Menke, 1976:335). Its natural range is uncertain, as its presence on some widely-dispersed
oceanic islands strongly suggests introduction by humans. Of the four species known from
the Philippines, this is certainly the commonest. I have found it present at most rural
localities, and at some it is among the most apparent of all solitary wasps.
Pison argentatum has been the subject of a number of investigations (Bordage, 1912;
Iwata, 1964a, b; Pagden, 1934; Williams, 1919; Yoshimoto, 1965; summarized by Iwata,
1976). In this paper I add to what is known through a more extensive treatment of nest
structure and prey choice and the ﬁrst data on primary sex-ratio.
Materials and Methods
This study was conducted sporadically from 1982–1985. Data are mainly from six
localities: Batangas – the marine research station of De La Salle University, at Matuod,
Lian, Batangas, Philippines, 148019N 1208329E, September–October 1985.
Baybay – the Visayas State College of Agriculture, near Baybay, Leyte, Philippines,
108459N 1248479E, at various times in 1982–1984.
Hilongos – Baas, Hilongos, Leyte, Philippines, 108229N 1248459E, October 1982.
Palawan – the Palawan National Agricultural College, near Aborlan, Palawan,
Philippines, 98269N 1188339E, April 1984.
Bukidnon – Busco, Quezon, Bukidnon, Philippines, 88169N 1248589E, December 1984.
Sabah – the Koh Bersatu Estate oil-palm plantation, 115 km west of Sandakan, Sabah
(Borneo), Malaysia, 58429N 1178099E, May 1985.
Dedicated with respect and affection to the memory of Howard Ensign Evans.
Accepted 21 May 2004; revised 27 July 2004
Ó2004 Kansas Entomological Society
JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY
77(4), 2004, pp. 565–572
The ﬁrst and last localities are separated by 8.32 degrees of latitude (918 km). Unless
otherwise stated, all data from a given locality were taken in the same season.
In determining the relative age of cells in active nests I compared brood age and took
note of any unﬁnished cell. Relative age of cells in old nests was determined by noting
which cell utilized another’s pre-existing wall.
Because I was interested in the ﬁnal size of unconstrained nests, I excluded all active
nests that had not clearly stopped growing, as well as any in which there was not clearly
room to add at least one cell beyond the last one, or in which I doubted that the substrate
could support additional weight.
In counting prey, I utilized not only fresh, paralyzed spiders from active nests but dried
fragmented ones from old nests. In the latter case, I determined and counted only
cephalothorax capsules, which are adequate to identify to family.
To determine offspring sex-ratio, I kept pupae from two localities for several weeks and
sexed all emerged adults. Where no adult emerged, I opened the pupal case to sex the
unemerged individual. Unless it had died as a prepupa or young pupa or had been
parasitized, this could usually be done with conﬁdence.
All sampling was haphazard. I collected all nests that I could ﬁnd and recorded what I
could of cell numbers, prey and offspring. In a given area I either censused thoroughly or
not at all, so that a bias toward larger, more conspicuous nests is unlikely in the nest size-
Specimens of P. argentatum collected C.K. Starr and identiﬁed by S.G. Reyes or C.K.
Starr in the collections of the University of the Philippines (Los Ban˜os), Visayas State
College of Agriculture, and De La Salle University (Manila) will serve as vouchers.
Results and Discussion
Nest Sites and Structure
By ‘‘nest’’ is meant a group of closely associated cells evidently made by a single wasp.
Nest contents indicated ages of brood that could have been produced by a single female in
a single nesting season. Nests at all localities were in sheltered situations. None was
exposed to rain, dripping water, or direct sunlight. This may explain why they were not
found on tree trunks or vertical rock faces. Bordage (1912) reported that P. argentatum
nests disintegrate if exposed to rain.
Pison argentatum nests are found most abundantly on buildings and other human-made
structures, in which case the substrate is usually a wall or shallow groove. Away from
buildings, they are found mostly on hanging roots under stream or road banks, the putative
situation in which the nest evolved.
Three basic nest types are known from Pison (Bohart and Menke, 1976; Evans et al.,
1980; Iwata, 1964b): a) free mud nests, b) pre-existing cavities, modiﬁed and closed with
mud, and c) burrows excavated in soil. Most P. argentatum nests are of the ﬁrst type, but
this species also uses cells of old mud nests built by other, larger wasps (Sceliphron and
Eumenes spp.) as cavities within which to build its own cells. At Bukidnon and Palawan, I
often found old cells of a medium-sized Eumenes sp. (Vespidae: Eumeninae), each with
two P. argentatum cells formed by means of a partition across the middle inside. The third
type is unknown in P. argentatum.
The basic unit of a P. argentatum nest is a spheroidal mud cell, about 10 mm in length
and about 9 mm in diameter, with walls about 1/2 mm thick. Where a cell joins the substrate
it is not lined, so that the substrate forms part of the inner walls. This includes the situation
566 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY
Fig. 1. Three nests at different stages of construction, on hanging roots under a road bank. Life size. a. Initial
cell, not yet fully provisioned or closed. b. Four cells, not yet plastered, the newest cell not yet fully provisioned
or closed. c. Completed nest, about ﬁve cells; plastering obscures cell outlines. The newest cell (at top) is much
less plastered than the others, though, suggesting an anomalous plastering after the fourth cell was closed, and
before the last one was begun. Scale bar ¼10 mm.
Fig. 2. Nest of 12 cells, to show age progresssion of cells. The nest was in a 90-degree vertical wall-angle in
a building; in the ﬁgure it is laid horizontal, with the original orientation indicated. Based on contents only,
progression from older to younger cells was as follows:
The relative ages of cells a–d are unknown according to this line of evidence, although all are older than eand all
others. Similarly, it is unknown whether cell jis older than ior/and k, but it is younger than all others except l.
Examination of shared cell-walls indicates that the full progression is almost certainly in alphabetic sequence.
Scale bar ¼10 mm.
VOLUME 77, ISSUE 4 567
where the substrate is formed by one or more existing cells, so that where they touch
adjacent cells always have a shared wall. Before a cell is provisioned, it is completed except
for a circular opening at the top about 3 mm across (Fig. 1a). This is closed only once, when
provisioning is ﬁnished.
The arrangement of cells in a nest shows no strongly consistent pattern, as expected
where each is functionally self-contained and no particular pattern is needed for structural
support. Just two general pattern rules are apparent:
1. Where the substrate is an approximately ﬂat surface or a groove on a ﬂat surface, the
cells tend to form a line (Fig. 2), with new cells added serially at one end only. There is
therefore a clear linear age-progression.
2. Where the arrangement departs clearly from the horizontal, new cells are added at the
top. I found no exception to this rule. It is most obvious in a nest that is still being built
or in which a cell was left unﬁnished. As in most hunting wasps, P. argentatum works
on one cell at a time, so that the age progression of cells is uncomplicated.
The status of rule 1 is uncertain when the nest is not based on a ﬂat surface but on
a hanging ﬁlament, as under a road bank. Of the 11 multi-cell nests in this situation found
at Hilongos, the cells of two were in a single vertical line, those of four were on either two
or three sides of the substrate root (Fig. 1c), and the rest simply presented clusters of cells
(illustrated by Williams, 1919).
As far as I am aware, all mud-building sphecids and eumenines follow rule 2, where it
can apply, for a readily apparent reason. The cell entrance is on the upper half of the cell—
in P. argentatum often at the very top—so that to build a new cell below an existing one
would be difﬁcult, even if the wasp has a place to stand while she builds.
Where the arrangement of cells is not strictly linear, rule 2 cannot be absolute.
Nonetheless, the trend is readily apparent even in nests at Hilongos that climbed their roots
two or three cells at a time or in a disorderly cluster. It is more apparent where the nest
consists of two or three columns of cells on a ﬂat surface, as exempliﬁed in Fig. 2.
As in most wasps, the nest does not have a ﬁxed size or number of cells. Iwata (1976)
notes maximum number of cells recorded for ﬁve Pison species, ranging from 4 in P.
chilense Spinola to 24 in P. argentatum. Figure 3 shows the distribution of cell number in
223 nests at Baybay and 38 at Bukidnon. Eleven nests at Hilongos each had 2–9 cells.
Mean cells per nest was 5.5 at Baybay, 3.2 at Bukidnon and 5.0 at Hilongos.
Economy of material is evidently a key factor in the evolution of wasp nests (Jeanne,
1975). Iwata (1976) states the plain rule that in all mud-daubers economy is increased by
nesting on a ﬂat surface, rather than on a root or stem, as cells are not lined. That is, the
substrate serves as a wall without elaboration. It is probably because of economy of
materials rather than structural support that P. argentatum, like many other mud-daubers,
shows a clear tendency to utilize inside-corners of buildings (Fig. 2) and others grooves,
As noted by Iwata (1976), in some species this tendency is extended to the habit of
renting, i.e., utilizing old cells of other species, with minimal alteration and addition of
new mud, instead of building entire new cells. It is not surprising, then, to ﬁnd P.
argentatum reutilizing old Eumenes sp. cells where these are abundant in the same nesting
situations. Occasional renting by habitually free-building Pison spp. may well be much
more common than is reported. It initially came to my own attention because I was
studying Eumenes at the same time.
568 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY
Iwata’s (1976) characterization of the walls between adjacent cells of Pison as
‘‘partitions’’ can be misleading. While such a shared wall is a partition from the point of
view of brood or a parasite inside a cell, to a female wasp it is rather a part of the substrate
for the new cell. It seems more useful, then, to view such walls not as dividers between cells
but as an element in economizing on nesting materials. The exact saving in material from
utilizing substrate depressions and grouping cells has not been determined, although it
appears to be considerable.
Figures 1c and 2 illustrate an additional element of nest structure. In Fig. 2 the borders
between cells are distinct, while in Fig. 1c the nest resembles an unstructured lump of earth,
so that it is difﬁcult from the outside to know where the cells join or even how many cells
there are. This is due to plastering, the application of loads of mud to the surface in addition
to that forming the cells proper. Iwata (1976) speaks of heavy plastering as the rule in P.
argentatum, but I have not found this to be the case among nests on buildings. At all study
localities a large fraction of such nests were bare, and at Baybay very few were plastered.
By contrast, plastering was the rule among nests on roots, as noted by Williams (1919).
The ﬁgures illustrate the extremes in plastering on ﬁnished nests. At intermediate stages,
plastering may consist of ﬁlling in the angle between cells, perhaps only at the nest-
substrate interface, laying low ridges leading away from the nest at the interface, and/or
laying low ridges on the nest surface.
The most probable function of plastering is to disguise the nest from parasites. Slight
plastering obscures the positions of individual cells, while more complete plastering may
Fig. 3. Number of cells in 223 completed nests at Baybay and 38 completed nests at Bukidnon.
VOLUME 77, ISSUE 4 569
obscure the very nature of the nest. In my own experience, it is often hard to recognize
a heavily plastered nest under a road bank among very similar-looking lumps of plain mud
on roots. It seems unlikely that plastering can physically stop a parasite capable of
penetrating cells proper, but it may signiﬁcantly increase the time and effort required to do
so. This, combined with uncertainty about which masses of dried mud are nests and which
are just solid lumps, may be enough to reduce parasitism. Certainly, parasite pressure on
mud-building wasps is important. All Southeast-Asian Pison,Sceliphron (Sphecidae) and
Eumenes (Vespidae) species whose nests I have examined in quantity show high rates of
brood-loss to parasites (unpubl. data).
I have the impression that nests under stream- and road-banks are more heavily
plastered than are those those on buildings. If so, this would support the hypothesis that
plastering is a facultatively expressed activity that functions to disguise nests in a setting
where other blobs of mud are present. However, an observation on nest structure in
Eumenes pyriformis (Fabr.) seems inconsistent with this interpretation. This common
species, which nests in much the same situations as P. argentatum in the Philippines, tends
also to group its cells in a linear series and shows similar variation in plastering. I have not
found any clear difference between nests on buildings and under banks in the usual amount
of plastering in this species or the less common E. fulvipennis (Smith).
An additional important feature of plastering is that it almost always occurs as a single
ﬁnal stage in nest-building, rather than piecemeal after completion of each cell. The
signiﬁcance of this may be expressed in two equivalent ways. First, it indicates a new level
in the hierarchical structure of the nesting sequence, above that of building and provisioning
the single cell. Second, it shows that a given nest has a particular number of cells and is not
just the sum of cells built until the wasp is somehow prevented from continuing.
Prey were almost all long-sighted hunting spiders of families in which most species do
not build webs (Table 1), mostly very small individuals or species. At all localities except
Sabah, jumping spiders (Salticidae) made up a majority of individuals counted. Web-
builders (all Araneidae) were the only other ecological class deﬁnitely represented,
accounting for 26/1087 (2%) of all prey examined from conﬁrmed P. argentatum cells.
Some apparent P. argentatum cells in old Eumenes cells at Palawan yielded short-
sighted hunters (Thomisidae) and web-builders (Araneidae and Theridiidae), but it is not
certain that all cells were provisioned by P. argentatum, and they are omitted from Table
1. These spider families are known as usual prey of other Pison spp. (Evans et al., 1980;
Iwata, 1976). It is curious that they showed up only in a group of closely-associated old
Eumenes cells—quite possibly utilized by a single Pison female—where they accounted
for 31/57 (54%) of prey.
Table 1. Numbers of prey spiders according to family from about 75 Pison argentatum cells at ﬁve localities
in the Philippines and Borneo.
Family Batangas Baybay Hilongos Palawan Sabah Total
Salticidae 71 211 363 93 9 747
Lycosidae 29 56 11 44 34 174
Oxyopidae 0 49 0 5 2 56
Araneidae 1 0 0 0 25 26
Total 101 316 374 142 70 1003
570 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY
Iwata (1976) recorded only salticids from P. argentatum cells, but it is seen here that the
wasp takes a broader range of long-sighted hunting spiders. Iwata recorded pisaurids as
prey of P. strandi, and I suspect that their absence from P. argentatum cells in this study is
because available species are much too large.
The sizeable taxonomic variation in prey among different localities is not surprising, and
seasonal differences within a locality are to be expected. However, it is plausible that the
greatest variation is between nests, as is suggested by the distribution of less common prey
types. For example, most of the 29 wolf spiders (Lycosidae) from Batangas came from
a single nest. If this reﬂects the true pattern, it suggests a strong learned component in prey
choice. Alternatively, it may simply reﬂect local abundance at the time of nesting.
Table 2 shows an individual prey-choice pattern whose signiﬁcance is unclear. This is
the complete record of prey from the nest shown in Fig. 2. It is almost as if the wasp were
choosing a ratio of about 1 oxyopid: 6 salticids, although it is hard to suggest any dietary
signiﬁcance in such a pattern.
The number of prey per cell varies considerably, and the complete range from my
samples is contained within the one nest just described, 7–21 per cell. This indicates
a great deal of variation in prey size, at least within the Salticidae. Iwata (1976) reports
a range of 3–19 prey/cell in P. argentatum.
It bears mention that active Pison cells would seem to be an excellent source of small
salticids for taxonomic study. In my experience, these include a good fraction of mature
The egg is laid obliquely on the abdomen, usually near the base and almost always on
the dorsolateral part. I have found eggs only in completely provisioned and closed cells,
indicating that the egg is laid on the last prey. These observations are consistent with what
has been reported for Pison spp. (Evans et al., 1980; Iwata, 1976).
Sex-ratio data from three samples from the Philippines (Table 3) are ambiguous. Each
sample is strongly biased, but together they give a balanced ratio. The size difference
Table 2. Numbers of prey spiders according to family in the eight provisioned cells of a Pison argentatum
nest from Baybay (see Fig. 2).
Totalef g h i j k l
Salticidae 7 6 18 13 16 12 9 15 96
Lycosidae 0 3000000 3
Oxyopidae 0 132213315
Total 7 10 21 15 18 13 12 18 114
Table 3. Sex of pupae and emerged adults from Pison argentatum nests at three localities in the Philippines.
Baybay (3 nests) Hilongos (4 nests) Batangas (1 nest) Total (8 nests)
Females 26 12 12 50
Males 17 29 5 51
Total 43 41 17 101
Females/Males 1.53 0.41 2.40 0.98
VOLUME 77, ISSUE 4 571
between adult females and males is very slight. In the absence of a larger data set, more
systematically collected, it seems reasonable to accept these data as corroborating the null
hypothesis of a balanced primary sex-ratio.
The nest structure and prey results largely corroborate what has already been written
about those species known to build free nests: P. argentatum,P. erythropus Kohl, P.
ignavum Turner, P. koreense (Radoszkowki), P. obliteratum F. Smith and P. ruﬁpes
Shuckard (Bohart and Menke, 1976; Evans et al., 1980; Iwata, 1976). However, present
results substantially extend and specify the variation known in P. argentatum.
I am grateful to Stephen G. Reyes for conﬁrming the identity of Pison argentatum,to
Juliet Can˜ete for ﬁeld assistance, and Allan Hook, Kunio Iwata and Robert Matthews for
criticism of an earlier version of this paper. The collecting trip to Hilongos was funded by
the Visayas State College of Agriculture. Thanks to Judith Acero and the Palawan
National Agricultural College for facilitating my stay in Palawan. In Bukidnon I was the
guest of Joe Zubiri and the Bukidnon Sugar Milling Company, and at the Sabah locality of
Vincent Au and the Koh Bersatu Estate Company.
This paper was drafted in the hospitable house of Humberto & Eunice Montes and Jesse
Casas in Borongan, Eastern Samar.
Baltazar, C. R. 1966. A catalogue of Philippine Hymenoptera (with a bibliography, 1758–1963). Paciﬁc Insects
Bohart, R. M., and A. S. Menke. 1976. Sphecid Wasps of the World. University of California Press, Berkeley.
Bordage, E. 1912. Notes biologiques recueillies a` l’ıˆledelaRe´union. Bulletin Scientiﬁque de France et Belgique
Evans, H. E., R. W. Matthews, and A. W. Hook. 1981 (1980). Notes on the nests and prey of six species of Pison
in Australia. Psyche 87:221–230.
Iwata, K. 1964a. Ethological notes of four Japanese species of Pison (Hymenoptera, Sphecidae). Mushi 38:1–6.
Iwata, K. 1964b. Bionomics of non-social wasps in Thailand. Nature and Life in Southeast Asia 3:323–383.
Iwata, K. 1976. Evolution of Instinct. Smithsonian Institution Press, Washington. 535 pp.
Jeanne, R. L. 1975. The adaptiveness of social wasp nest architecture. Quarterly Review of Biology 50:267–287.
Menke, A. S. 1988. Pison in the New World: a revision (Hymenoptera: Sphecidae: Trypoxylini). Contributions of
the American Entomological Institute 24(3):1–171.
Pagden, H. T. 1934. Biological notes on some Malayan Aculeate Hymenoptera I. (Sphecoidea and Vespoidea).
Journal of the Federated Malay States Museum 17:458–486.
Williams, F. X. 1919. Philippine wasps studies. Part II. Descriptions on new species and life history studies.
Hawaiian Sugar Planters’ Association Experiment Station Bulletin (Entomological Series) 14:19–186.
Yoshimoto, C. M. 1965. Nesting activity of the mud-daubing wasp, Pison argentatum Shuckard in Hawaii
(Hymenoptera, Sphecidae, Trypoxyloninae). Paciﬁc Insects 7:291–294.
572 JOURNAL OF THE KANSAS ENTOMOLOGICAL SOCIETY