CORRESPONDENCE zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 5 France, R.L. (1996) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Vegetafio 124, 67-72 6 France, R. and Steedman, R. Trans. Am. Fish. Sot. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA (in press) 7 Malej, A., Faganeli, J. and Pezdic, J. (1993) Mar. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA The eusociality continuum revisited mating and territorial encounters, even in taxa that are otherwise strictly solitary, so that virtually all species would be labelled eusocial. Biol. 116,565-570 8 Schindler, DE. et al. (1998) in Food Webs: integration of zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Pat t er ns and Dynamics (Polis, G. and Winemiller, K.O., eds), pp. 96-108, Chapman & Hall 9 Meyer, J.L., Schultz, E.T. and Helfman, G.S. (1983) Science 220,1047-1049 Reply from F. Boero et zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA al. Costa and Fitzgerald1 reviewed recent reconsiderations of the definition of eusociality. They concluded that all previous definitions are flawed, primarily because they (apparently) exclude interesting social taxa. As an alternative, Costa and Fitzgerald argued that the presence of a communication system is the hallmark of (eu?)sociality. We disagree with their analyses and conclusions for three reasons. Costa and Fitzgerald also appear to approve of the idea (attributed to Wicslo7) that it is best not to categorize societies at all, but rather to simply note and describe each taxon as a phylogenetically unique entity. However, this approach would block the fundamental scientific goal of seeking common causal principles underlying convergent phenomena, in this case, social systems with strikingly similar features. Therefore, we reiterate and expand our suggestion that social systems can be usefully arrayed along a continuum with reproductive skew as the metric. France’s approach reflects a minimalistic examination of the functioning of life, a fashionable attitude that has reached its apex in a statement by Maddox1 in one of his last editorials for Nature: ‘Life is chemistry’. When we say that there is a strong functional connection between plankton and benthos, we do not speak only about jellyfish, we speak also about phytoplankton and copepods (the core of plankton) as having benthic stages that ‘flow’ to the plankton again, exporting biomass and (presumably) energy. These issues are overlooked by chemically-oriented researchers, and it is obvious that you will never find what you never look for, especially if you are biased by standardized techniques. France says: ‘with the exception of organisms such as filter-feeding moliuscs or deposit-feeding profundal macroinvertebrates, there is little mixing between the C flows of pelagic and benthic food webs...’ Sponges, cnidarians, most annelids, bryozoans, kamptozoans, hemichordates, crinoids and many ophiuroids, and tunicates, together with filter-feeding molluscs or deposit-feeding macroinvertebrates are not exceptions. They are the rule! And they mostly feed from the water column. Which consumers are considered by France? Echinoids, macrocrustaceans, gastropods? Most of them have planktonic larvae. And these larvae feed, and are themselves food, in the plankton. First, Costa and Fitzgerald misconstrued the fundamental purpose of the new definition of eusociality proposed by Sherman et a/.2. Our goal was not to find a ‘common denominator’ of all social systems, but rather to develop and quantify the notion of a ‘reproductive division of labor’, which is central to understanding one of the great evolutionary problems, one which had its origins in Darwin’s” puzzlement over physical castes of sterile workers and altruism in social insects: how can reproductive self-sacrifice be reconciled with natural selection? Sherman et al. suggested quantifying reproductive division of labor as reproductive skew. An important consequence of using skew is that both vertebrate and invertebrate systems can be arrayed along one continuum, paving the way for revealing cross-taxonomic comparisons of the causes and consequences of reproductive partitioning in these societies. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONM H. Kern Reeve Paul W. Sherman Section of Neurobiology and Behavior, Cornell University, Ithaca, NY 14853-2702, USA (hkrl@cornell.edu) Laurent Keller Contrary to Costa and Fitzgerald’s implication, the eusociality continuum is not a continuum of interestingness (and ‘eusociality’ is not a label for ‘really interesting’), and it does not exclude groups ‘characterized by non-reproductive cooperative interactions’ (i.e. groups within which there is no offspring care, such as tent caterpillar associations), as Costa and Fitzgerald assumed. In fact, the continuum (free of the assumption of overlap in generations) does include such groups (i.e. those with no offspring care) since the skew in lifetime reproductive success (LRS) is affected by any cooperative behavior that has future reproductive consequences, such as reducing survivorship of some group members. Thus, the eusociality continuum and reproductive skew will illuminate the patterns of altruism in tent caterpillars and any other taxa in which cooperative behavior affects lifetime reproduction. Institut de Zoologie et d’Ecologie Animale, Universite de Lausanne, Bltiment de Biologie, 1015 Lausanne, Switzerland, and Zoologisches Institut, Bern University, Ethologische Station Hasli, Wohlenstrasse 50a, 3032 Hinterkappelen, Switzerland (Lkeller@ulys.unil.ch) zyxwvutsrqponmlkjihgfedc References Costa, J.T. and Fitzgerald, T.D. (1996) Trends Ecol. Evol. 11, 285-289 Sherman, P.W. et al. (1995) Behav. Ecol. 6, 102-108 Darwin, C. (1859) The Origin ofspecies, John Murray Crespi, J. and Yanega, D. (1995) Behav. Ecol. 6, 109-115 To consider one part of a cycle (or one part of a community) and taking it as the complete representation of the studied system is classical for chemically oriented ecological research. When results are transferred to graphs they even look interesting, but the biology is lost. Life becomes chemistry. We dispute this approach, and we are not alone (see, for instance, Ref. 2). Our article, and this brief postscript, are perhaps too extreme and crude, but they are a reaction to an extreme (and crudely prevailing) attitude that sees biology reduced to chemistry. Keller, L. and Krieger, M.J.B. (1996) Nature 380, 208-209 Keller, L. and Perrin, N. (1995) Proc. R. SOC. London Ser. B 260,311-315 Wicslo, W.T. in Social Competition and Cooperation in insects and Arachnids: /I Evolution ofSociality(Choe, J.C. and Crespi, B.J., eds), Cambridge University Press (in press) Ferdinand0 Boero Genuario Belmonte Dipartimento di Biologia dell’Universita di Lecce, 73100 Lecce, Italy Giovanni Fanelli Stefano Piraino Fernando Rubino Second, Costa and Fitzgerald (following Crespi and Yanega4) criticized the continuum idea because the reproductive skew measure is upwardly biased by chance differences in reproduction. Thus, they confused the practical issue of the best statistical measure of skew with the deeper question of whether it is useful to array societies along a continuum. Concerning the practical issue, the authors failed to mention that Keller and KriegerS and Keller and Perrin6 have modified the skew index to account for chance variation in LRS. Reply from J.T. Costa and T.D. Fitzgerald Istituto Sperimentale Talassografico - CNR, Via Roma 3, 74100 Taranto, Italy References 1 Maddox, J. (1995) Nature 378,435-438 2 Threlkeld, S.T. (1994) Hydrobiologia 275/276, 293-300 Finally, Costa and Fitzgerald’s ‘solution’ to the eusociality definition debate is apparently simply to lump together as (eu?)social all taxa in which there is communication. This fails to address the evolutionary problem that stimulated our continuum idea. Moreover, it also enlarges the concept of eusociality so greatly as to render it nearly useless as a classificatory tool. For example, communication is involved in almost all We find it puzzling that Reeve et a/. state here (but not elsewhere) that non-reproductive social groups are not excluded from their eusocial continuum. Alloparental care is an explicit defining characteristic of their concept of eusocialityl. Indeed, these authors appear to be confusing a eusociality continuum with a continuum of reproductive skew. Reproductive skew exists in all populations, even those in which there is no social interaction, and an array of species based solely on lifetime reproductive success (LRS) is irrelevant to the question of social evolution. If indeed they intend to drop the criterion of alloparental care, their eusociality concept becomes, in 472 0 1996,Elsevier Science Ltd zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA TREE vol. II, no. II November 1996 View publication stats View publication stats