Larval orientation in scorpions: phylogenetic patterns and ecological speculations.

Warren E. Savary

Presented as a poster at the
American Arachnological Society meetings
Tucson, Arizona
August 1996

     Female scorpions carry their larvae on their dorsa in one of three patterns of orientation: random, longitudinal, or transverse. Randomly oriented larvae (Figures 1, 2) are stacked in two or more layers on the dorsum of the mother and show no group consistency in alignment. Transversely oriented larvae (not illustrated) are stacked two to three layers deep, are aligned at right angles relative to the longitudinal axis of the mother, face medially, and leave a narrow gap along the mother's midline (Francke 1982). Longitudinally oriented larvae (Figures 3, 4) are borne in a single layer, are aligned parallel to the longitudinal axis of the mother, face anteriorly, and typically have their prosomas directed downward and their metasomas curled so that the aculeus points towards their mother's dorsum (see also Stahnke 1944, Williams 1969). In the longitudinal pattern of orientation, each larva is in direct contact with its mother's cuticle.

     The published literature on scorpions contains, through direct description or photographic depiction, information on the larval orientation of at least 21 species of scorpions belonging to six families. I have been able to expand this body of information through observation of captive scorpions and unpublished photographs to include 25 species of scorpions in six families (see Table 1). This information, while admittedly representing a small minority of scorpion species, nonetheless allows for the evaluation of character polarity for the various orientation patterns, supports an independent hypothesis of common ancestry for three genera of vaejovid scorpions, and invites the advancement of hypotheses of adaptation in explanation of the derived orientations.

Character polarity and phylogenetic significance of larval orientation patterns:

     Random orientation of larvae occurs in all six of the families for which information is available and is the exclusive pattern of orientation known for four of those families (Buthidae, Euscorpiidae, Ischnuridae, and Scorpionidae). Transverse orientation has been reported only from the diplocentrid genus Diplocentrus Peters (Francke 1982), while longitudinal orientation appears restricted to the vaejovid genera Vaejovis Koch, Syntropis Kraepelin and Serradigitus Stahnke (see Table 1). When the known distribution of larval orientation patterns is viewed in the context of the most recently proposed hypotheses of scorpion phylogeny (Lourenco 1985, Stockwell 1988, 1990, 1996), the polarity of the orientation patterns becomes apparent (see Figs. 5-8). In each of the phylogenetic hypotheses, the nearest outgroup to the Diplocentridae (Scorpionidae in Lourenco's 1985 hypothesis and Stockwell's 1990 and 1996 hypotheses, and the Scorpionidae + Ischnuridae lineage in Stockwell's 1988 hypothesis) exhibits random orientation. The outgroup comparison method for establishing character polarity (Maddison, Donoghue and Maddison 1984) clearly establishes transverse orientation as apomorhic. Similarly, the nearest outgroups to the Vaejovidae for which larval orientation is known (Chactidae sensu latus in Lourenco's 1985 hypothesis, the Euscorpiinae + Megaconninae lineage in Stockwell's 1988 hypothesis, Euscorpiinae in Stockwell's 1990 hypothesis, and the Diplocentridae + Scorpionidae + Ischnuridae lineage in Stockwell's 1996 hypothesis) exhibits random orientation, either exclusively (to the extent known) or as it's plesiomorphic state, but does not exhibit longitudinal orientation. Outgroup comparison thus also establishes longitudinal orientation as apomorhic. In both cases, further reference to secondary and tertiary outgroups further strengthens the assesment of transverse and longitudinal orientations as derived.

     To date, transverse orientation has been reported only from a single species, Diplocentrus whitei (Gervais), and thus sheds little light on the phylogenetic relationships within the genus Diplocentrus or within the family Diplocentridae. The recorded occurrences of longitudinal orientation are more plentiful and diverse, and, as previously suggested by Williams and Savary (1991), unite the Vaejovis minimus group (Pseudouroctonus bogerti and P. minimus groups sensu Stockwell 1996) with Syntropis, Serradigitus and other members of the genus Vaejovis (notably excepting the type species, V. mexicanus, for which no data are available). This association is corroborated by at least one morphological synapomorphy: in those taxa which exhibit longitudinal orientation of the larvae, the lamellar hooks of the spermatophore are elevated along, and adnate to, the lamella. In other vaejovid taxa, the lamellar hooks are distinctly basal in position and, in some, are well separated from the lamella. Such an interpretation is not inconsistent with the most recently suggested phylogeny of the Vaejovidae (Stockwell 1996), and would appear to resolve polychotomies therein (see Figure 9).

Possible adaptive significance:

     The assessment of longitudinal orientation as a derived state allows its evolution to be viewed from an adaptational perspective (see Coddington 1988). Several adaptive advantages have been proposed for the transport of larvae by adult scorpions. These include protection against predation (Vannini, Ugolini, and Marucelli 1978; Vannini, Ugolini, and Carmignani 1985; Vannini, Balzi, Becciolini, Carmignani, and Ugolini 1985), continuous selection by the mother of the optimal microclimate for the larvae (Williams 1969; Vannini, Ugolini, and Marucelli 1978; Ugolini, Giannellini, and Vannini 1981; Vannini, Ugolini and Carmignani 1985), and some sort of trophic exchange between mother and larvae (Vannini, Ugolini, and Marucelli 1978), involving either food (Alexander 1977), wax, or water (Vannini, Ugolini, and Carmignani 1985; Vannini, Balzi, Becciolini, Carmignani, and Ugolini 1985). Hjelle (1974) has suggested that it is necessary for larvae to ascend the mother's back in order to undergo a successful molt.

      Laboratory studies have shown that larvae disassociated from the mother have a lower survival rate than those which remain on the mother's back (Hjelle 1974; Vannini, Ugolini, and Marucelli 1978; Francke 1979; Vannini and Ugolini 1980, Vannini, Ugolini, and Carmignani 1985). Predation was clearly not a factor in these studies, and its role relative to larval survival in natural environments has yet to be reported. Vannini, Ugolini, and Marucelli (1978) noted that larval Euscorpius italicus raised on the mother survived and molted, while larvae raised on pieces of wood or dried models under the same conditions died before molting. Ugolini, Giannellini and Vannini (1981) raised Euscorpius larvae both in isolation and in association with mothers at relative humidities of 10%, 60%, and 90%. Larvae raised with their mothers at all three humidities and larvae raised in isolation at 90% relative humidity lived longer than larvae raised in isolation at 10% or 60% relative humidity. This would suggest that association with the mother can enhance larval survival in inhospitable microclimates and that continuous microclimatic selection by the mother may not be necessary.

     Larvae do not feed. Second instars (nymphs) capture and consume their own prey. They have been observed feeding on the remains of their mother's prey (Williams 1969; Schultze 1927), but do not seem dependent upon maternally acquired food and may feed wholly on their own even when confined with the mother (Vannini, Ugolini, and Marucelli 1978). Alexander (1977) has suggested that food sharing and help in locating food are advantages that accrue from the mother/young association. Williams (1969) noted that young scorpions begin dispersing as second instars soon after the cuticle hardens and suggested that matemal feeding is insignificant.

     Larvae of Euscorpius flavicaudus kept on females treated with tritiated water showed a highly significant relationship between emitted radiation and time spent on the female's back (Vannini, Ugolini, and Carmignani 1985; Vannini, Balzi, Becciolini, Carmignani, and Ugolini 1985). Vannini and his co-workers suggested that tritium could have been transferred to the larvae as liquid water or vapor, or been incorporated into the female's epicuticular lipids and transferred to the larvae by contact. Either type of transfer (wax exchange or water exchange) could account for enhanced survival rates of maternally associated larvae in low humidities. Vannini and his co-workers suggested that the slow rate of cuticular lipid turnover made water exchange the more likely explanation. Hadley (1974) found that newly molted scorpions experience higher rates of water loss due to cuticular transpiration than do intermolt specimens with tanned cuticles. It is likely that the soft-cuticled, untanned larvae are also subject to significant water loss through cuticular transpiration.

     Proposed explanations of the adaptive values for the longitudinal orientation of larvae center on the structural security of the arrangement. No similar proposals have been made for the transverse orientation of larvae. Williams (1969) suggested that longitudinal orientation allows more young to be efficiently carried per unit space, allows firmer positioning for each member of the litter, and minimizes accidental fall to the ground. Haradon (1972) suggested that individual attachments offer a more secure riding position. It is of interest to note, however, that both Williams (1969) and Haradon (1972) have reported reduced activity in larvae-carrying females of the genus Vaejovis which bore longitudinally oriented larvae. Williams (1969) noted that female Vaejovis remain inactive within shelters while carrying larvae. Haradon (1972) noted that captive female Vaejovis with larvae remain sedentary. I have encountered a single larvae-carrying female of Serradigitus gertschi (Stahnke) in the field. It was active on the ground surface and attempted to retreat at my near approach. Active wandering by scorpions bearing randomly oriented larvae has been previously noted. Adult Euscorpius spp. (Chactidae) with young on their backs have been observed to wander and feed occasionally (Vannini, Ugolini, and Marucelli 1978) and adult Centruroides exilicauda (Buthidae) are frequently seen on the ground surface while carrying young (Williams 1969). I have also observed a larvae-carrying captive female Centruroides gracilis (Latrielle) engaging in courtship within 24 hours of giving birth.

     Another hypothesis that attempts to explain the adaptive value for the longitudinal pattern of orientation can be derived from the above-cited studies conducted by Vaninni, Ugolini and their colleagues. One of the characteristics of the longitudinal pattern of orientation is the presence of a single layer of larvae, which contrasts with the multiple layers found in the other known orientation patterns. In the longitudinal pattern, each larva remains in direct contact with the mother's cuticle. In large litters, larvae maintain this contact by tilting the prosoma downwards to contact the mother's cuticle, thus giving the larval assemblage the appearance of leaning books on a bookshelf. If water is indeed transferred from mother to larvae via the cuticle, then such continuous direct contact would clearly be of value where the possibility of dessication is increased. As the members of the lineage exhibiting longitudinal orientation are primarily associated with the North American deserts, where dessication is a risk, this speculation likely merits testing.

Literature cited:

Alexander, A. J. 1977. Parental behaviour in scorpions. XVth International Ethological Conference, Bielfeld. 23-31 August 1977. Abstracts (Contributed Papers) - 2.

Angermann, H. 1957. Uber verhalten, spermatophorenbildung und sinnesphysiologie von Euscorpius italicus Hbst. und verwandten arten (Scorpiones, Chactidae). Zeitschrift Tierpsycol. 14: 276-302.

Baerg, W. J. 196 1. Scorpions: biology and effect of their venom. Agric. Exper. Sta. Bull. 649, Div. Agric., Univ. of Arkansas, Fayetteville: 1-34.

Coddington, J. A. 1988. Cladistic tests of adaptational hypotheses. Cladistics 4(l): 1-20. Francke, 0. F. 1976. Observations of the life history of Uroctonus mordax Thorell (Scorpionida, Vaejovidae). Bull. British Arachnological Society 3(9): 254-260.

Francke, 0. F. 1979. Observations on the reproductive biology and life history of Megacormus gertschi Diaz (Scorpiones: Chactidae: Megacorrninae). J. Arachnology 7: 223-230.

Francke, 0. F. 1982. Birth behavior in Diplocentrus bigbendensis Stahnke (Scorpiones, Diplocentridae). J. Arachnology 10: 157-164.

Francke, 0. F., and S. K. Jones. 1982. The life history of Centruroides gracilis (Scorpiones, Buthidae). J. Arachnology 10: 223-239.

Hadley, N. F. 1974. Adaptational biology of desert scorpions. J. Arachnology 2: 11-23.

Haradon, R. M. 1972. Birth behavior of the scorpion Uroctonus mordax Thorell (Vaejovidae). Entomol. News 83: 218-221.

Hjelle, J. T. 1974. Observations on the birth and post-birth behavior of Syntropis macrura Kraepelin (Scorpionida: Vaejovidae). J. Arachnology 1: 221-227.

Lourenco, W. R. 1985. Essai d'interpretation de la distribution du genre Opisthacanthus (Arachnida, Scorpiones, Ischnuridae) dans les regions Neotropical et Afrotropical. Etude taxonomique, biogeographique, evolutive et ecologique. Doctoral dissertation, Universite Paris VI (Pierre et Marie Curie), 287 pp.

Maddison, W. P., M. J. Donoghue, and D. R. Maddison. 1984. Outgroup analysis and parsimony. Systematic Zoology 33(l): 83-103.

Probst, P. 1967. Der geburtsvorgang beim skorpion Isometrus maculatus. Revue Suisse Zool. 74(3): 616-619.

Rosin, R., and A. Shulov. 1963. Studies on the scorpion Nebo heirochonticus. Proc. Zool. Soc. London 140(4): 547-575.

Sissom, W. D., and 0. F. Francke. 1983. Post-birth development of Vaejovis bilineatus Pocock (Scorpiones: Vaejovidae). J. Arachnology 111: 69-75.

Stahnke, H. L. 1944. Do baby scorpions eat mother? Nature Magazine 37(10): 521-522. Stahnke, H. L. 1966. Some aspects of scorpion behavior. Bull. So. Calif. Acad. Sci. 65(2): 65-80.

Stockwell S. A. 1988. A preliminary phylogenetic analysis of scorpion families. American Arachnology 38: 9.

Stockwell, S. A. 1990. Revision of the Phylogeny and Higher Classification of Scorpions (Chelicerata). Ph.D. Dissertation (1989), University of California, Berkeley. University Microfilms International, Ann Arbor, Michigan. 413 pp.

Stockwell, S. A. 1996. Scorpion Phylogeny World Wide Web site at URL  "".

Ugolini, A., V. Giannellini, and M. Vannini. 198 1. Adaptive value of parental care in Euscorpius (Scorpiones). Monitore Zool. Ital. (N. S.) 15: 327.

Vachon M., R. Roy, and M.Condamin. 1970. Le development post-embryonaire du scorpion Pandinus gambiensis Pocock. Bull. Inst. Fond. d'Afr. Noire 32(2): 412-432.

Vannini, M., A. Ugolini, and C. Marucelli. 1978. Notes on the mother-young relationship in some Euscorpius (Scorpiones: Chactidae). Behav. Ecol. Sociobiol. 7: 45-47.

Vannini, A. Ugolini, and 1. Cannignani. 1985. Mother-young relationships in Euscorpius (Scorpiones): Trophic exchange between mother and larvae. Monitore Zool. Ital. (N.S.) 19: 172.

Vannini, M., M. Balzi, A. Becciolini, 1. Carmignani, and A. Ugolini. 1985. Water exchange between mother and larvae in scorpions. Experientia 41 (12): 1620-162 1.

Williams, S. C. 1969. Birth activities of some North American scorpions. Proc. California Academy of Sciences 37(l): 1-24.

Williams, S. C. 1971. Birth behavior in the South African scorpion Hadogenes. Pan-Pacific Entomologist 47(l ): 79-80.

Williams, S. C. and W. E. Savary. 199 1. Uroctonites, a new genus of scorpion form western North America (Scorpiones: Vaejovidae). Pan-Pacific Entomologist 67(4): 272-287.

Click here to return to my home page.