Jaumal of the Lepidopterists' Society 56(2), 2002, 90-97

DEW-DRINKING BY MALE MONARCH BUTTEHFLIES, DANA US PLEXIPPUS (L.)

DENNIS FREY, ROBERT ROMAN AND LINDSAY MESSETT Biological Sciences Department, California Polytechnic State University, San Luis Obispo, California 93407, USA

ABSTRACT. We examined the early morning activity of overwintering monarch butterflies at a central coast California site during their period of "mass-mating" from mid-January to mid-March in 1998 and 1999. The first year of the study took place during El Nino weather when regional precipitation was approximately 50% greater than normal. The second year occurred during La Nina conditions with approximately 15% less precipitation than normal. On "rain-free" mornings many of the butterflies left aggregations and flew to an adjacent meadow where they landed on grass and drank from dew droplets. Males were much more likely to make morning meadow visits than females and such males had 6.7% higher body moisture than males that remained in the clusters. The rate of male meadow use increased seasonally both years and males were frequently seen attempting to mate follOwing dew-drinking meadow visits. Early morning meadow-males resembled males at­ tempting to mate in the morning more than males that were not mating, i.e., that remained in aggregations. Meadow-males were smaller, had relatively smaller dry-weight abdomen mass, and had more wing damage in addition to their greater moisture content. Variation in local solar radiation, wind speed and evapotranspiration during the previous 24 h was positively associated with early morning meadow use by both males and females. Male mating effort during the previous day was pOSitively associated with early morning dew-drinking by males but not females. Three hypotheses regarding dew-drinking activity are considered. The results are consistent with two perspectives: dew drinking results from dehydrating activity such as courtship and spermatophore transfer during the previous day and/or it represents strategic male behavior in an­ ticipation of "need" for future reproductive effort. Additional key words: Danaus plexippus, dew-drinking, mating effort, dehydration, El Nino, La Nina.

Each fall monarch butterflies, Danaus plexippus Overwintering at California sites on the other hand oc­ (L.), migrate from an extensive late summer breeding curs during the wet season, yet desiccation is a prob­ range in North America to geographically restricted lem for monarchs. MOisture-producing storms de­ overwintering habitats (Brower & Malcolm 1991). crease in frequency during February and March, Monarchs from eastern breeding populations overwin­ occur sporadically, last for only brief periods and are ter in several mountainous areas dominated by Oya­ often followed by many days of dry conditions. For ex­ mel fir forests near Mexico City (Urquhart & Urquhart ample, during the wet 1998 El Nino event, on 58 days 1978, Brower 1985, Calvert & Brower 1986). Those in January, February and March, moisture loss originating west of the Rocky Mountains overwinter at through evapotranspiration exceeded moisture gain wooded sites located along the Pacific Ocean between from precipitation along the central coast of California Bolinas, California and Enseneda, Mexico (Sakai & (data source: California Department of Water Re­ Calvert 1991, Lane 1993). sources, California Irrigation Management Informa­ The general eco-region macroclimate where these tion System from weather stations located in sites are located, along with local topography and veg­ Guadalupe, California). This potential for desiccation etation, prOvide microclimate conditions favorable for was even more pronounced during the dry 1999 La monarch overwintering (Brower et al. 1977, Calvert et Nina winter. al. 1982). Conditions thought to be important for over­ An intense period of mating occurs between late wintering include: (1) generally cool temperatures that January and early March at both Mexico and Califor­ minimize the rate of butterfly activity and rate of body nia overwintering sites during which males compete fat utilization (Chaplin & Wells 1982, Masters et al. vigorously for mating opportunities (Hill et al. 1976, 1988, Alonso-Mejia et al. 1997), (2) a pattern of daily Tuskes & Brower 1978, Van Hook 1996, Frey et al. minimum temperature that reduces the likelihood of 1998). During a lengthy copulation the male transfers lethal freezing (Alonso-Mejia et al. 1997), (3) enough a spermatophore to his partner's reproductive tract, moisture to minimize desiccation (Calvert & Lawton which may equal 10% of his body weight (Oberhauser 1993), (4) low wind velodty to minimize disruption 1988). Monarch spermatophores contain a small am­ and desiccation (Leong et al. 1991), and (5) exposure of poule of spermatozoa along with various nutrients, but butterflies in cluster formations to brief mosaics of di­ consists primarily of water (up to 93%) and females rect solar radiation to enable periodic thermoregula­ may use this "transferred" water for somatic mainte­ tory basking (Calvert & Brower 1986, Frey et al. 1992). nance (Oberhauser 1992, Pham 1997). At California Overwintering at Mexico sites occurs during the re­ overwintering sites mating could increase water deficit gional dry season but dehydrating conditions are re­ for males due to the high water content of sper­ duced somewhat because these sites occur at high ele­ matophores and the seasonal increase in evapotranspi­ vations (2400 to 3600 m) where cloud forests provide ration, solar radiation and temperature. Overwintering relatively high humidity and closed-canopy forests re­ sites in Mexico and California are usually associated sult in wind abatement (Calvert & Brower 1986). with a nearby source of water (Calvert & Brower 1986, 92 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY

TABLE 1. Regional weather variables recorded during 1998 and from the rest of the body and weighed separately. Dry 1999 on (a) the morning of meadow counts and (b) the day preced­ weight of the abdomen relative to total dry body ing meadow counts. Values are daily means ± SE (standard error). weight was used as an index of body condition or an Variablel 1998 1999 p2 approximate "fat-content" index.

(a) Temperature of initial flight 11.2 ± 0.4 10.4 ± 0.3 0.106 Estimation of male mating effort. We estimated (OC) male mating effort in two ways. In the first technique Change in temperature 2.2 ± 0.4 4.4 ± 0.4 < 0.001 we videotaped courtship responses of males to pinned (OC )3 monarch specimens positioned on platforms in areas (b) ± ± Evapotranspiration (mm) 1.9 0.3 2.2 0.2 0.459 of high mating activity. From late morning recording Solar radiation (W m-2) 144 ± 17 162 ± 10 0.649 Wind velocity (ms-I ) 2.7 ± 0.2 2.5 ± 0.1 0.344 sessions (55 min duration) we summed the number of "visits" by males to the models as well as the number 1 Evapotranspiration, solar radiation, and wind velocity values were down-loaded from CIMIS data bases for stations #52 in San of times males landed on them and tried to couple Luis Obispo, California and #120 in Guadalupe, California main­ with the models. For more details of this approach see tained by tbe California Department of Water Resources. Falco (1999). These values were converted to standard 2 P-values are from Mann-Whitney tests. normal scores and used as an index of population-level 3 Change in temperature was recorded daily on-site as the differ­ male mating effort. The second technique involved ence in temperature of the initial morning flight from cluster trees versus the temperature during the second morning transect count late morning counts (approXimately 1 h duration) of approximately 45 minutes later. "ground pairs", i.e., a male attempting to couple with a female, in the area where mating attempts frequently mornings during 1999 warmed more qUickly between occurred. We followed the protocol of Frey (1999) and transect counts than during 1998 (Table 1). converted the number of observed attempts per Moisture content and condition of monarchs. minute to standard normal scores. Males were captured weekly during 1998 from three different categories: (1) dew drinkers: males located in RESULTS the meadow with their probosces extended into dew Entry rates into the meadow. A few monarchs droplets on grass, (2) mating pairs: males that had just were occaSionally found in the meadow during count captured and coupled with a female near the meadow, # 1 before the beginning of morning flight activity. (3) aggregating individuals: males captured with a net They were usually wet with dew, much more so than attached to a long-reach pole from the aggregations ones collected later in the day, and females were often one hour after ambient temperature exceeded flight found near males. We used an abdominal palpation threshold. Males from the meadow and those mating technique (Van Hook 1999) and found that most of the were captured immediately following the second tran­ females had a large detectable spermatophore present. sect count of monarchs in the meadow. During 1999, This suggested that they had mated recently and pos­ males were captured in the meadow as described for Sibly had spent part of the previous night in copula in category 1 above on the mornings of 23 February and the meadow rather than more typically being attached 2 March. Males were captured from clusters as de­ to males that perched in the canopy of nearby trees. scribed for category 3 above on 20 February and 27 During the first year of the study the Pismo Beach February 1999. Mating males were not sampled dur­ monarch population declined from a peak abundance ing 1999. of 125,000 butterflies in late December 1997 to less The number of wings that had membrane tears or than 1000 individuals by March 1998. During the sec­ portions missing were counted for each butterfly. We ond year the population declined from 100,000 mon­ placed each male individually in a freezer proof zip­ archs in December 1998 to less than 1000 butterflies lock bag and stored them in an ice-chest until they in March 1999, In both years the sex ratio became in­ were transported to our lab where they were frozen at creaSingly male biased seasonally which is a pattern -20°C. These specimens were later weighed (wet previously reported for this site (Frey & Leong 1993, weight) and then dehydrated in a drying oven at 60°C 1995, Frey et al. 1998). We computed an index of sex­ for 40 h. Individuals were weighed immediately on re­ speCific increase in meadow abundance to adjust for moval from the freezer to minimize bias due to accu­ the follOwing factors: (1) differences in abundance and mulation of water that condenses on thawing speci­ sex ratio pattern between years, (2) seasonal declines mens. FollOwing dehydration their dried body weight in abundance due to dispersal that began in January was recorded and moisture content was estimated as each year, and (3) slight day-to-day differences in the the difference between wet and dry weight. FollOwing time interval between transect counts. This index was dry weight measurement, the abdomen was severed computed by dividing the number of new butterflies 94 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY

TABLE 4. Results of Spearman correlation tests of weather variables as predictors of meadow entry rate during the 46 sampling dates of the 1998 and 1999 monarch mating seasons combined; (a) tests based on temperatures recorded on-site during the morning of meadow counts as predictor variables, (b) tests based on weather conditions during the day preceding meadow counts as predictors.

Influence on male meadow entry rate Influence on female meadow entry rate Predictor variable Spearman "z" P Spearman "z" P (a) Temperature of initial flights (OC ) -1.47 0.142 -0.87 0.386 Change in temperature during morning (OC) 1.58 0.114 1.66 0.097 (b) Evapotranspiration (mm) 1.92 0.055 1.41 0.159 Solar radiation (W m-2) 2.85 0.004 2.34 0.019 Wind velocity (m S-l) 1.74 0.082 0.561 0.576

the following day and vice versa. Female meadow-use perature during January, February, and March ofl998 was unrelated to the previous day's mating effort approximated the 3D-average for the Pismo Beach area (Spearman Z = 0.19; N = 20, P = 0.85). but were 4%, 3%, and 9% below normal respectively Female per capita use of the meadow did not during 1999. change Significantly during the course of the season There are several reasons why water may be partic­ (Spearman correlation; Z = 1.36, P = 0.18, N = 46). On ularly important to males during the mass mating pe­ the other hand, male per capita entry into the meadow riod at overwintering sites beyond a general need to increased Significantly as the mass mating season pro­ reduce negative systemic phYSiological effects that gressed (Spearman correlation; Z = 2.80, P = 0.005, N may accompany desiccation. Three hypotheses re­ = 46). Relatively greater proportions of the male popu­ garding early morning dew-drinking are given in Table lation made early morning visits to the meadow later in 5 along with speCific testable predictions for each the season. hypothesis. Early morning temperatures could influ­ ence flights from cluster trees (Table 5, Current DISCUSSION Weather-hypothesis 1) because flight thresholds can Male and female monarchs flew from overnight vary slightly among individuals in overwintering popu­ roosting aggregations to a nearby meadow as morning lations (Masters et al. 1988, Anson Lui pers. com.). Al­ temperature warmed above the £light threshold. Indi­ ternatively, or in addition, weather associated with de­ viduals of both sexes were observed drinking from dew hydration during the previous day could influence droplets on grass blades, i.e., they had their probosces dew-drinking on the follOwing morning (Table 5, Pre­ extended into the droplets for long periods of time. vious Day Weather-hypothesis 2). To test these first Males collected from the meadow had 6.7% greater two hypotheses we combined data from both years of moisture content than males collected from aggrega­ the study for the follOwing three reasons: (1) per capita tions, which also suggests that they had been drinking. meadow entry rate did not differ between years for ei­ Leong et al. (1992) reported monarch water loss be­ ther sex (Table 2), (2) on-site temperature during the tween 6% and 10% per day when held without water initial morning flights did not differ between years, under lab conditions (19.PC and 44.9% relative hu­ and (3) the regional evapotranspiration, solar radia­ midity). Males were much more likely to visit the tion, and wind velocity on days preceding our observa­ meadow than females during the early morning when tions of dew-drinking did not differ between the two dew was available (Table 2). years (Table 1). Dew-drinking patterns were similar between the Neither temperature-related prediction of the "cur­ two years of this study even though the amount of pre­ rent weather constraint" hypothesis, i.e., Table 5-hy­ cipitation differed greatly (Table 2). This pattern indi­ pothesis 1, was borne out in this study (Table 2 & cates that monarch moisture levels are closely regu­ Table 4a). Early morning conditions, however, can in­ lated and influenced more by short-term events (e.g., fluence meadow use since monarchs are constrained the past 24 to 48 h) than over a longer term. In con­ from flying at temperatures below approximately lOOC trast, year-to-year differences in relative abdomen and they seldom fly under overcast conditions (pers. mass (Table 3, larger abdomens in 1999) probably re­ obs.). One of the two predictions of the "previous day flected overall differences in temperature regimes be­ weather" hypothesis, i.e., Table 5-hypothesis 2, was tween EI Nino and La Nina seasons and the fact that confirmed. Levels of solar radiation and other weather fat reserves are used more rapidly under higher tem­ variables associated with dehydration on a given day peratures (Chaplin & Wells 1982). Average daily tem- were positively correlated with dew-drinking rate on 96 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY

al. 1999; but see Scriber 1987). Puddling is often ence of temperature on crawling, shivering, and flying in over­ viewed as differential foraging, which either supplies wintering monarch butterflies in Mexico, pp. 309-314. In S. Malcolm & M. Zalucki (eds.), Biology and conservation of the ionic or nutrient resources needed by males (Arms et monarch butterRy. Los Angeles, Science Series No. 83, Natural al. 1974) or is a means for males to acquire resources History Museum of Los Angeles County. that they transfer to females during copulation (Scul­ ALONSO-MEJIA, A., E. RENDON-SALINAS, E. MON TES INOS-PATINO & L. BROWER. 1997. Use of lipid reserves by monarch butterRies ley & Boggs 1996). Alternatively, or in addition, male overwintering in Mexico: implications for conservation. Eco!. biased puddling in Lepidoptera could provide mois­ Applic.7:934-947. ture for reproductive activities, as well as, mainte­ ARMS, K., P. FEENY & R. LEDERHOUSE. 1974. Sodium: stimulus for puddling behaviour by tiger swallowtail butterRies. Science nance, as outlined above for monarchs. Monarchs also 185:372-374. puddle at Mexico overwintering sites (pers. obs.) but BAILEY, R. 1978. Descriptions of the ecoregions of the United puddling has not been reported for monarchs in west­ States. U.S.D.A. Forest Service Intermountain Region, Ogden Utah. ern North America. BECK, J. , E. MUHLENBERG & K. FIEDLER. 1999. Mud-puddling be­ Nectaring from nearby flowering plants has been havior in tropical butterRies: in search of proteins or minerals? observed for monarchs during the mass mating phase Oecologia 119:140-148. BELL, E., L. BROWER, W. CALVERT. J. DAYTON, D. FREY, K. LEON G, at a Mexico overwintering site (Alonso-Mejia et al. D. MURPHY, R. PYLE , K. SNOW & S. WEI SS. 1993. The 1997). Flower-visiting monarchs in Mexico were monarch project's conservation and management guidelines for smaller (dry mass), had Significantly less fat reserves preserving the monarch butterRy migration and monarch over­ wintering habitat in California. Portland, Xerces Society publi­ and had relatively greater moisture levels than cluster­ cations. captured ones (58% vs. 54% respectively for moisture BOGGS, C. & L. JACKSON. 1991. Mud puddling by butterflies is not as a proportion of wet body weight; our calculations of a simple matter. Eco!. Entomo!' 16: 123-127. BROWER, L., w. CALVERT, L. HEDRICK & J. CHRISTIAN. 1977. Bio­ data from Table 2 in Alonso-Mejia et al. 1997). The logical observations on an overwintering colony of monarch pattern for moisture content of these Mexico mon­ butterflies (Danaus plexippus, Danaidae) in Mexico. J. Lepid. archs is similar to those from our dew-drinking study Soc. 31:232-242. BROWER, L. & S. MALCOLM. 1991. Animal migrations: endangered and indicates that nectaring can also provide monarchs phenomena. Amer. Zool. 31:265-276. with a source of water. Few nectar sources however BROWER, L. P. 1985. New perspectives on the migration biology of occur near the Pismo overwintering habitat and mon­ the monarch butterRy, Danaus plexippus L., pp. 748-785. In M. A. Rankin (ed.), Migration: mechanisms and adaptive signifi­ archs are seldom observed nectaring during January cance. University of Texas, Austin, Texas. and February (pers. obs.). Likewise, they seldom drink CALVERT, W. & L. BROWER. 1986. The location of monarch butter­ from open water that usually occurs in shaded areas of fly (Danaus plexippus L.) overwintering colonies in Mexico in relation to topography and climate. J. Lepid. Soc. 40: 164-187. the habitat, suggesting that monarchs probably obtain CALVERT, W. & R. LAWTON. 1993. Comparative phenology of varia­ most of their water from dew or water droplets. tion in size, weight, and water content of eastern North Ameri­ Dew drinking by monarchs, like nectaring and pud­ can monarch butterflies at five overwintering sites in Mexico, pp. 299-307. In S. Malcolm & M. Zalucki (eds.), Biology and dling, may be "triggered" by a number of proximate­ conservatiun uf the monarch butterRy. Los Angeles: Science Se­ level factors and may provide water that is used in sev­ ries No. 83, Natural History Museum of Los Angeles County. eral functional contexts. Early morning dew drinking CALVERT, W., W. ZUCHOWSKI & L. BROWER. 1982. The impact of forest thinning on microclimate in monarch butterRy (Danaus by monarchs seems to be a response to short-term de­ plexippus L. ) overwintering sites in Mexico. Boletin de la So­ hydration as well as strategic behavior associated with ciedad Botanica de Mexico 42:11-18. male reproductive activity. CHAPLIN, S. & P WELLS. 1982. Energy reserves and metabolic ex­ penditures of monarch butterRies overwintering in southern California. Eco!. Entomo!' 7:249-256. ACKNOWLEDGEMENTS FALCO, L. 1999. Variation in male courtship behaviors of the This work was supported by a California State Faculty Support monarch butterfly (Danaus plexippus L.) at central California grant to Dennis Frey. We are grateful to California Parks and Recre­ overwintering sites. M.S. Thesis, California PolytechniC State ation Department for allOwing us to conduct monarch research on University. their property. Monarchs were collected under a permit issued to FREY, D. 1999. Hesistance to mating by female monarch butter­ Dennis Frey by the California Parks and Recreation Department. flies , pp. 79-87. In J. Hoth, L. Merino, K. Oberhauser, 1. SpeCial thanks go to Lincoln Brower and Lisa Falco who provided Pisanty, S. Price & T. Wilkinson (eds.), 1997 North American inSight and suggestions regarding both the project and manuscript. Conference on the Monarch ButterRy. Commission for Envi­ We are also thankful for valuable reviews by Karen Oberhauser, ronmental Cooperation, Montreal. Carla Penz, and an anonymous reviewer. FREY, D. & K. LEONG. 1993. Can microhabitat selection or differ­ ences in 'catch ability' explain male-biased sex ratios in overwin­ tering populations of monarch butterRies? Anim. Behav. LITERATURE CITED 45:1025-1027. ADLER, P. 1982. Soil- and puddle-visiting habits of moths. J. Lepid. --. 1995. Reply to Nylin, Wickman & Wiklund regarding sex Soc. 36:161- 173. ratios of California overwintering monarch butterflies. Anim. ADLER, P. & D. PEARSON. 1982. Why do male butterRies visit mud Behav. 49:515-518. puddles? Canadian J. of Zoo!. 60:322-325. FREY, D., K. LEONG, D. FREDERICKS & S. RASKOWITZ. 1992. Clus­ ALONSO-MEJIA, A., J. GLENDINNING & L. BROWER. 1993. The influ- ter patterns of monarch butterflies (Danaus plexippus (L.) GENERAL NOTES

Journal of the Lepidopterists' Society 56(2), 2002, 104-106

IMMATURE STAGES OF SAIS ROSALIA (NYMPHALIDAE, ITHOMIINAE) Additional key words: life-history, Mechanitini, Solanaceae,

The immature stages of butterflies in the subfamily chewing small holes in the blade. Although solitary, Ithomiinae (Nymphalidae) are relatively well known larvae were not cannibalistic; several ins tars could be for most genera, with good descriptions available in reared together without losses. The caterpillars rested the literature (DeVries 1987, Brown & Freitas 1994 in a J-shaped position on the underside of the leaves. and references therein), However, information is When disturbed, caterpillars dropped off the leaf, sus­ scarce or absent for many small genera (such as pending themselves by silk threads. Roswellia, Eutresis, Athyrtis, Paititia, Aremfoxia, Vela­ Egg (Fig. la). White, elongated, slightly pointed at the apex, with dyris, Velamysta, Dygoris, and Hypomenitis), and still 15- 18 longitudinal ridges and 9-12 transverse ridges (similar to that described in Brown & Freitas 1994). Duration 3-4 days. incomplete for most larger diversified genera (Hya­ Larvae. First instar. White, turning green after first meal; legs, lyris, Hypothyris, Napeogenes, Hyposcada, Oleria, prolegs and anal plate black. Head black; average width 0.48 mm [thomia, Pteronymia, Greta, Hypoleria). In the tribe (SD = 0.03, n = 11). In dorsal view, the lateral tubercles (present in Mechanitini, there is adequate information for most all instars) could be observed easily as small rounded projections on species and all genera (Brown & Freitas 1994), except each abdominal segment. Maximum length 4 mm. Duration 2-3 days. Second and third instars. Dark green to leaden gray, with a Forbestra (larva briefly described in Drummond 1976) white collar on the prothorax and a lateral series of short yellow tu­ and Sais (minimal information on the egg and larva in bercles along the abdomen; legs, prolegs and anal plate pale. Head Table 1 and Fig. 2 of Brown & Freitas 1994, from rear­ black; average width (second ins tar) 0.75 mm (S D = 0.03, n = 12), ing in Goiania, Goias, Brazil, in 1968). In this paper, all (third) 1.19 mm (SD = 0.04, n = 12). Maximum length 8 mm (sec­ immature stages of Sais rosalia (Cramer, 1779) are de­ ond) or 13 mm (third instar). Duration 3 days (second) or 8-9 days (third). Fourth (final) instar (Fig. Ib, c). Dorsum dark gray, ventral scribed and illustrated. region pale gray, with a white collar on the prothorax and a lateral se­ Sais rosalia (near subspecies rosalinde Weymer, ries of short yellow tubercles (tubercles on AI, A2, A7 and A8 more 1890 =paraensis Haensch, 1905, see Lamas 1994) was developed); legs and prolegs dark. Head blaek or rarely brown; aver­ studied on banks and islands of the Teles Pires River, age width 1.74 mm (SD = 0.04, n = 10). Maximum length 25 mm. north of Alta Floresta, state of Mato Grosso, Brazil, in Duration 5-6 days. The pre-pupa loses the contrasting color-pattern and becomes reddish, adopting an arched configuration (Fig. Id). February and June 2000. On the latter visit, many eggs Pupa (Fig. Ie-g). Elongated, slightly arched ventrally (about and larvae were collected on a solanaceous vine, prob­ 60°) between the second and third abdominal segments, beige with ably of the genus Lycianthes (very similar to the host a general gold reflection and black stripes on the wing cases (rarely recorded in Goiania; a live plant has been kept to await dark with reduced reflective areas); cremaster red. Ocular caps short flowering). Additional eggs (n = 19) were also obtained and pOinted. Length 15 mIll. Duration 13 days (n = 12). from a wild caught female kept in a plastic bag with The immature stages of Sais rosalia are similar to the host plant. The larvae were kept with leaves of the those of other Mechanitini, including egg shape and host plant in plastic boxes that were cleaned daily. Egg size, number of longitudinal and transverse ridges, a size was measured as height and width; and the larval lateral series of tubercles on the abdominal segments of head capsule size as the distance between the most ex­ the larvae, and the pupa elongated, slightly arched and t~rnal ocelli (as in Freitas 1991, 1993); all capsules reflective (Motta 1989, Brown & Freitas 1994). Con­ were retained for confirmation of growth stages. trary to Mechanitis, which has long pointed lateral tu­ Adults, preserved larvae, capsules, and pupal skins are bercles, the larva of Sais bears short lateral tubercles, a in the collection of the first author. feature shared with Thyridia, Forbestra and Scada About 20 eggs and four first ins tars were collected in (Brown & Freitas 1994). Perhaps the most interesting the field on several food plant individuals growing near feature in the life history of the studied population of the riverbanks in sunny places. The plants varied from Sais rosalia is that there were only four larval stadia. 50 em to 2 m high, and had small soft leaves. The iso­ All known ithomiine larvae have at least five ins tars (six lated eggs and larvae were found on the underside of instars were observed once in Placidina, AVLF pers. mature leaves. Females were observed ovipositing in obs.), with the exception of Tellervo zoilus (Fabricius, the late afternoon, from 1500 to 1800 h. After inspect­ 1775), also with four instars (Ackery 1987:272, with A. ing the plant, the female landed on the upper side of a G. Orr); this Australian species has been placed in a mature leaf and curved the abdomen to lay an isolated separate subfamily by some authors (Ackery 1987). egg on the underside. After hatching, caterpillars ate Only four instars were also observed in 12 larvae of part of the eggshell, and later began to eat the leaves, Sais rosalia brasiliensis (reared to pupae on the Teles 106 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY

(Vit6ria Da Riva Carvalho and her son Edson), for extensive and Thesis, University of Florida, xvi + 361 pp. patient logistic support, housing and food during work in this re­ FREITAS, A. V 1. 1991. Varia9ao morfol6gica, ciclo de vida e sis­ gion in February and June 2000. Gerardo Lamas, Annette Aiello tematica de Tegosa claudina (Eschscholtz) (Lepidoptera, and Carla Penz made valuable comments on the manuscript. This Nymphalidae, Melitaeinae) no estado de Sao Paulo, Brasil. research was partly supported by fellowships from the Brazilian Revta. bras. Ent. 35:301-306. Conselho Nacional de Desenvolvimento Cientffico e Tecnol6gico --. 1993. Biology and population dynamics of Placidula eu­ (CNPq) and the Funda9ao de Amparo it Pesquisa do Estado de ryanassa, a relict ithomiine butterfly (Lepidoptera: Ithomiinae). Sao Paulo (BIOTA/FAPESP program, grants 98/05101-8 and J. Lepid. Soc. 47:87-105. 00101484-1 ). LAMAS, G. 1994. Los Danainae e Ithomiinae descIitos por R. Haensch (Lepidoptera: Nymphalidae). Shilap 22:271- 297. LITERATURE CITED MOTTA, P. C. 1989. Analise filogenetica de Ithomiinae (Lep.: Nymphalidae) com base nos ovos: Rela9aO com plantas hos­ ACKERY, P. R. 1987. The danaid genus Telleroo (Lepidoptera: pedeiras. MS Thesis. Universidade Estadual de Campinas. Nymphalidae)-a cladistic approach. Zool. J. Linn. Soc. Campinas, SP, Brazil. 89:273-294. BROWN JR., K. S. & A. V L. FREITAS. 1994. Juvenile stages of ANDRE V 1. FREITAS AND KEITH S. BROWN JR., Museu de lthomiinae: overview and systematics (Lepidoptera: Nymphali­ Hist6ria Natural and Departamento de Zoologia, Instituto de Biolo­ dae). Trop. Lepid. 5:9-20. gia, Universidade Estadual de Campinas, C.P 6109, Campinas, Sao DEVRIES, P. J. 1987. The butterflies of Costa Rica and their natural Paulo, 13083-970, Brazil. history. Papilionidae, Pieridae, Nymphalidae. Princeton Univ. Press, Princeton, New Jersey. 327 pp. DRUMMOND III, B. A. 1976. Comparative ecology and mimetic re­ Received for publication 20 August 2001; revised and accepted 26 lationships of Ithomiine butterflies in Eastern Ecuador. Ph.D. November 2001.

Journal of the Lepidopterists Society 56(2),2002, 106- 108

MATINGS WITHOUT SPERMATOPHORE TRANSFER AND WITH TRANSFER OF TWO SPERMATOPHORES IN CALLOPHRYS XAMI (LYCAENIDAE) Additional key words: spermatophore production, copulation, sexual selection.

In Lepidoptera, males normally transfer one sper­ tivity during a longer period (in several species there is matophore during copulation (Drummond 1984). How­ a negative correlation between female receptivity and ever, some studies indicate that in some matings no the degree of distention of the corpus bursa) (Drum­ spermatophore is transferred (although this does not mond 1984) or if permits the transfer of more sperm necessarily means that the female is not inseminated; (for example, if spermatophores can contain only a cer­ Drummond 1984), There are several possible explana­ tain maximum amount of sperm). However, the transfer tions for this fact: (a) exhaustion of substances necessary of multiple spermatophores may be disadvantageous for for building spermatophores as a result of frequent mat­ many species, since the last spermatophore needs to be ing (Drummond 1984); (b) male or female disabilities, at least partially digested before re-mating because such as deformations in the genitalia or in the reproduc­ sperm migration to the spermatheca reqUires proper tive tract resulting from disease or defective develop­ alignment of the spermatophore tube with the ductus ment; or (c) mate choice (rejection) after initiation of seminalis and this alignment is more difficult in the copulation by females (i.e., females somehow inhibit or presence of another spermatophore (Drummond 1984, prevent the transfer of spermatophores by certain Simmons & Siva-Jothy 1998). Thus, an alternative hy­ males; Eberhard 1996) or by males (i.e" males avoid to pothesis is that the transfer of more than one sper­ transfer spermatophores to certain females), Mate re­ matophore in one copulation is result of a male disabil­ jection could be achieved by interrupting copulations ity. before successful spermatophore transfer; in this case The multiplicity of explanations, and the theoretical copulations are expected to be of short duration relevance of many of them, indicates that to report mat­ (Cordero 1993). ings in which no spermatophore is transferred and in On the other hand, it has been found that sometimes which multiple spermatophores are transferred, as well males transfer more than one spermatophore in one as its possible causes, is important. During the course of copulation (Drummond 1984), This type of mating may three laboratory experiments on spermatophore pro­ be a male adaptation to sperm competition if the trans­ duction by males of the lycaenid butterfly Callophrys fer of multiple spermatophores decreases female recep- xami (Reakirt) (Cordero 1998), in which I observed 199