Associations Between Flies and Spiders: Bibiocommensalism and Dipsoparasitism?
Psyche 84(2):150-157, 1977.
This article at Hindawi Publishing: https://doi.org/10.1155/1977/26019
CEC's scan of this article: http://psyche.entclub.org/pdf/84/84-150.pdf, 1776K
This landing page: http://psyche.entclub.org/84/84-150.html
The following unprocessed text is extracted automatically from the PDF file, and is likely to be both incomplete and full of errors. Please consult the PDF file for the complete article.
ASSOCIATIONS BETWEEN FLIES AND SPIDERS:
BIBIOCOMMENSALISM AND DIPSOPARASITISM?*
MICHAEL H. ROBINSON AND BARBARA ROBINSON Smithsonian Tropical Research Institute
P.O. Box 2072, Balboa, Canal Zone, Panama There are numerous records in the arachnological and entomo- logical literature of relationships between spiders and flies other than the simple case of predator and prey. Bristowe (1941:362-370) reviewed a number of cases of parasitism and commensalism. Flies of the superfamily Drosophiloidea are involved in a number of more or less complex relationships with spiders. Chloropids parasitize spiders' egg cocoons and may actually perch on adult spiders (Bris- towe, 1941:367) while milichiids share food with spiders (Richards, 1953). McMillan (1975) has recorded an association between mili- chiid flies of the genus Desmometopa and two species of large Aus- tralian orb-weaving spiders. The flies moved about the host web and fed on prey items as they were being consumed by the spiders. In addition, the milichiid moved onto the host and apparently cleaned the mouthparts and anal region of the spider. McMillan does not state whether the flies remained on the spiders when they were not actively cleaning them nor does he state how many flies were present on the spider at any one time. We here report on several different associations between flies and spiders, all of which are commensal (in the broadest sense). We found milichiids asso- ciated with the golden-web spider Nephila cluvipes, unidentified flies were found as commensals of Argiope suvignyi, and chloropid flies were found in a similar relationship with Argiope argentata. All these relationships were discovered in Panama. We describe a case of milichiid commensalism with a predatory hemipteran and suggest that the complex relationship between Nephila and the milichiids may have evolved from such a relatively simple stage. We think that the term commensalism is not sufficiently specific to describe some of the relationships reviewed here and suggest two possible additions to the terminology of symbioses. 'Manuscript received by the editor October 31, 1977.
19771 Robinson & Robinson - Flies and Spiders 151 4J
Nephila clavipes and flies of the genus PhyUomyza We first found flies associated with Nephila clavipes in January 1976, on Barro Colorado Island, Canal Zone, Panama. We sub-
sequently found similar flies associated with this spider at a number of localities in the Canal Zone and elsewhere in Panama. Flies from four adult female N. clavipes at four different sites were col- lected. They were identified (see acknowledgments) as belonging to the genus Phyllomyza and all belonging to the same (unde- termined) species. All eleven insects were females. The flies rest on the dorsal surface of the spider and usually aggregate on the cephalothorax. Figure 1 shows eight flies resting on this area. The flies remain on the spider for long periods of time and are virtually inactive. When we set out to determine what the flies were doing sitting on the body of the spider we ran into a major practical problem. The Drosophila-sized flies were really too small to observe with the unaided eye. This problem was solved by adapting a stereo-binocular microscope for horizontal use, mounted on a camera tripod (Robinson & Smythe, 1976). With this device, under field conditions we could watch the insects under 10X or 20X magnification. It became apparent that on the cepha- lothorax of the spider the flies were not doing anything other than grooming themselves, sporadically shifting position and occasion- ally defecating. The bodies looked entirely normal and there was no evidence of oviposition or of penetrative feeding on the spider itself. (The mouthparts of milichiids could clearly not be used for piercing the spider's cuticle and sucking its internal fluids, but at this stage we did not know what the flies were.) Eventually we decided to feed the spider. This went through all phases of its predatory behavior without disturbance to the flies. Prey capture involved rushing out to attack the prey, biting it, wrapping it in silk, removing it from the web, transporting it back to the hub and there wrapping it once again before hanging it and feeding (details in Robinson & Robinson, 1973, for Nephila maculata ap- ply broadly to N. clavipes). This predatory sequence involves a great deal of violent movement, in space and of the spider's legs, throughout which the flies simply sat tight. At the hub the spider passed secretions into the prey and after about eight minutes the whole surface of the insect was covered in a film of liquid. At this stage the flies left the body of the spider and clustered on the surface of its prey. There they could be seen
19771 Robinson & Robinson - Flies and Spiders 153 dabbing at the liquid with their extended mouthparts. They quickly became swollen with food and their abdomens in particular were distended and almost spherical. The intersegmental membranes became very clearly visible and extended. After feeding the flies returned to the spider's cephalothorax. This timing of movement onto the prey item to coincide with its liquifaction seems to be fairly precise. It occurred in five out of five instances in which we pro- vided the spider with prey and watched the whole process from its inception. The spider on which we made these observations dis- appeared after four days and we replaced it (on the still-intact web) with an adult female Argiope argentata. Two flies settled on this spider and fed once on her prey before disappearing. (It is note- worthy that the Argiope was able to locate and successfully attack prey on the structurally very different Nephila web. This has pro- vided us with a useful tool for further studies of araneid predatory behavior.)
We saw very few cases where the spider reacted to the presence of the flies. Araneids seem to make very few responses to the ac- tivities of their larger kleptoparasitic associates, the theridiids of the genus Argyroides (see Robinson & Olazarri, 1971:34-5; Robin- son & Robinson, 1973:32).
Argiope savignyi and unidentified flies
While carrying out observations on Argiope savignyi in an in- sectary at Curundu, Canal Zone, Panama, the prey of two separate adult female spiders was visited by flies that did not alight on the spider at any stage. The flies "appeared from nowhere" and fed on liquifying prey items from which the spider was simultaneously feeding. On one occasion the spider's prey was a pentatomid and the entire insectary in which we were working was flooded with the penetrating odor of the hemipteran's defensive secretion. In the second case (Figure 2), the prey was a moth. In both instances the flies alighted on the spider's prey and never moved onto the spider at any stage. After feeding they simply flew off. We were unable to catch the three flies involved.
Argiope argentata and Conioscinella sp.
While censusing Argiope argentata along the Old Gamboa Road, Summit, Canal Zone, we found an adult female of this species con- suming a half-digested acridiid. On this prey were two small flies
154 Psyche [June
Arqwpv w~i;.,'tn'i feeding on a moth, t ~ o Hies (marked) arc visible on the prey.
that we succeeded in collecting. They were identified as chloropids of the genus Conio'icinetta. Several species of chloropids are known to parasitize the egg cocoons of spiders (see, for instance, Bristowe, 1941 :366-7).
Milichiids and a reduvild
At the La Fortuna dam site, Chiriqui. Panama, one of us (M H R) observed a reduviid Zefus trimacuhtus Distant, feeding on a sting-
19771 Robinson & Robinson - Flies and Spiders 155 less bee (Trigona cupira Sm.) which was impaled on the bug's pro- boscis. Around the prey item was a large number of small flies all apparently feeding. Eight of these flies were captured, many es- caped. The flies were identified as Neophyllomyza sp. The relationship between Phyllomyza sp. and its host Nephila clavipes is one that involves prolonged contact between the two species. We suspect that the same flies may remain on the spider for days at a time, leaving only to make very short feeding forays. Our attempts to paint-mark the tiny flies failed utterly so we cannot be certain on this point. In any case, the association seems to us to be distinctly more specialized than that described by McMillan (1975) for Desmometopa sp. Of course, this is a matter of inter- pretation. However, it is possible to suggest an evolutionary path- way from commensalism without contact (the reduviid and Argiope savignyi associates) through commensalism plus feeding excursions onto the host, to commensalism with sustained non-trophic con- tact. If the flies were cued into food sources by olfactory stimuli, as seems possible, then the pathway would involve a reduction of the detection distance. The strategy of waiting on the host must involve some interesting mechanism that allows the fly to "evalu- ate" the odds on food being available within its own feeding time- scale. At some stage the fly may be faced with "deciding'' whether to remain with a spider on the off-chance that it will catch food or using its remaining food reserves to fly off in search of another (more successful?) host. In this respect, the fly may be at an ad- vantage over the kleptoparasitic theridiid spiders that also asso- ciate with N. clavipes. It can probably range over greater distances, more quickly, in search of a new host than can the spiders. At least three species of theridiids associate with the golden-web spi- der; at least one of these regularly shares the host's meal, at the hub of the web (Vollrath, in press). Such kleptoparasitic spiders are small, but differ from the milichiids in having mouthparts ca- pable of penetrating insect cuticle.
Spiders of the genus Nephila may be particularly suitable as hosts for this kind of associate, They are large, build very efficient webs that are operated 24 hours per day, and show a considerable degree of web-site tenacity. The other large, diurnal, orb-web spi-
ders in Panama build much more ephemeral webs that are more susceptible to damage (they are nothing like as strong). They prob- ably spend much less time on their webs at any one site, a situation which may be less favorable to the development of a protracted association. The fact that the only time we have seen milichiids resting on an Argiope was when this spider had been placed on a Nephila web could indicate that the flies respond to some char- acteristic of the web in finding their hosts. On a recent trip to Papua New Guinea (May 1977), one of us (MHR) looked at over 500 adult Nephila maculata, hoping to find flies resting on the spi- der. None were found. Bristowe (1941:369) reports that R. N. Champion Jones saw a small fly crawl over the palps of Nephila maculata in India. Conceivably this is a case where a less sustained association has evolved.
As far as the milichiids are concerned it is at least possible that their relationship with Nephila may be more than a trophic one. The fact that all the Phflom~iza sp. that we collected were females is disturbing. The fact that we found Conioscinella feeding on the prey of Argiope argentata makes it possible that the egg parasites of araneids could also be commensals. The reverse could be true. There is some problem about finding terms that accurately de- scribe the relationship of the milichiids to their host(s). They are clearly commensals (in the broad sense) since they "share a table" with their hosts. However, they rely on the host liquifying the prey, they drink alongside the host and could perhaps be called bibiocommensals. The presence of fairly large numbers of flies feeding on a prey item could reduce the amount of food available to the host in a significant way, in which case the term parasite would be justifiable. A drinking parasite would be a dipsoparasite and this term is (to us) more euphonious than bibiocommensal. We thank the Insect Identification and Beneficial Insect Intro- duction Institute, USDA, Beltsville, Maryland (Chairman, Dr. L. Knutson), for identifying all the insects involved in this study. The experts involved were: C. W. Sabrosky, J. L. Herring, and S. W. Batra. Dr. Sabrosky drew our attention to a number of published works on the relationship between milichiids and spiders; this ad- ditional help is greatly appreciated.
Robinson & Robinson - Flies and Spiders
BRISTOWE, W. S.: The comity of spiders. Volume 2. Ray Society, London 1941 MCMII-LAN, R. P.: Observations on flies of the family Milichiidae cleaning Araneus and Nephila spiders. West Australian Naturalist, 13: 96 (1965) RICHARDS, 0. W.: On commensalism of Desmometopa with predacious insects and spiders. Proc. roy. ent. Soc. (Lond.) Ser. C. 18: 55-56 (1953) ROBINSON. M. H. AND OI.AZARRI, J.: Units of behavior and complex sequences in the predatory behavior of Argiope argenlata (Fabricius): (Araneae: Araneidae). Smith. Contr. Zool. 65: 1-36 (1971) ROBINSON, M. H. AND ROBINSON, B.: The ecology and behavior of the giant wood spider Nephila marculata (Fabricius) in New Guinea. Smith. Contr. Zool. 149: 1-76 (1973)
ROBINSON, M. H. AND SMYTHE, N.: A technique for observing the behavior of small animals under field conditions. Psyche 83: 210-212 (1976)
Volume 84 table of contents