1B. glandulosa is considered to be a cleistogamous form of P. cuspidatum.
²Earliest time at which oviposition was observed.

Upon reaching an acceptable acanth, females alighted and rotated their bodies (either direction) in a circle on the upper surface of the leaf, touching their abdomens down many times, and beating their wings slowly and rhythmically (1-2 times per second). They performed this “dance” on the leaf surfaces for up to a minute on any one leaf, or for several minutes of rotating on several adjacent leaves. While based only on qualitative observation, this behavior seemed to be related to the available leaf surface area on the potential hosts: if the leaf surface area was relatively large (e.g., Pseuderanthemum: 4-8 sq cm), the female remained on one leaf surface, but if the plant leaves were small (e.g., Justicia valerii: 1-3 sq cm), she moved between a few adjacent leaves and danced on them for several minutes before selecting or rejecting the plant as an oviposition site. Quantitative data is needed to test this hypothesis.

Once a leaf was selected, the female gradually stopped beating her wings while curling her abdomen onto the lower surface of the leaf, and began to oviposit. In all nine observed cases, the head and the front walking legs remained over the dorsal leaf surface. Except for the abdominal movements necessary for oviposition, the Anthanassa remained still, unless disturbed—in which case, rhythmic and then more rapid wing beats followed (based upon one instance where I inadvertently disturbed an ovipositing female). The earliest observed ovipositions started at 1220 hr (A. ardys) and 1149 hr (A. tulcis), and the latest began at 1319 hr (A. ardys) and 1153 hr (A. tulcis) (Table 1). Oviposition lasted 6-15 minutes in A. ardys, and 19-20 minutes in A. tulcis (Table 1). Clusters of 28-67 (A. ardys) and 73-96 (A. tulcis) eggs were laid; on one occasion two large clusters and one singly laid egg (A. ardys) were found on one leaf. It is not known if the clusters were from the same or different females. Eggs of A. adys were greenish white and bullet-shaped, with muted surface sculpturing. Eggs were 0.55mm in diameter and 0.6-0.7mm in height. Eggs of A. tulcis were yellow-green, spherical to bullet-shaped, and smooth with barely visible scaling on the surface. The eggs were 0.5mm in diameter and 0.4-0.5mm in height. There was no significant correlation between the duration of oviposition and the number of eggs deposited (R = 0.467, P = 0.35, Table 1—this regression was performed for A. ardys only). When finished ovipositing, females returned to the upper surface of the leaf and beat their wings slowly, basking for up to five minutes before flying away.

S. epaphus:

Females exhibited searching behavior similar to that of Anthanassa. When a female alighted on an acceptable oviposition site, she lowered her abdomen to the leaf surfaces a few times, and then extended her abdomen to lay a single egg over the course of 10-20 seconds. Females stopped beating their wings for only a few seconds while ovipositing. On the only observed host, Blechum pyramidatum, eggs were placed in curled leaves, the junctions of leaves and stems, or between flower bracts. Sometimes, a female returned to the same plant to deposit another egg. The eggs are spherical, 2 mm in diameter, and green with strong, vertical ribbing.

Anartia fatima:

The behavior of this butterfly was similar to that of S. epaphus. Single eggs (blue-green, spherical, approximately 1 mm in diameter) were laid on leaf surfaces or between flower bracts on Blechum, as observed at one lower-elevation site. Females also oviposited on leaf surfaces of low-growing Hydrocotyle sp. (Apiaceae) and Spermacoce assurgens (Rubiaceae), growing near small patches of Blechum. The five ovipositions observed lasted from 5-10 seconds each. On 22 and 24 March 1997, Feldman observed ovipositing females of Anartia fatima at La Selva Research Station in Heredia, Costa Rica, and found that females would land on the host (Blechum browneii) and take off again, subsequently ovipositing on the first plant she encountered (based on five observations of one female, one observation of another). Oviposition sites included Blechum browneii, Hydrocotyle mexicana (Apiaceae), a fern, and a dead leaf—grasses were not used as oviposition sites by these individuals.

Rearing Results

A total of 14 egg clusters were collected from 20 January to 1 February (Table 2). Six of these came from observed ovipositions of A. ardys, two were from observed ovipositions of A. tulcis, and four were found by searching acanth leaves (Table 2). All of the individuals in egg clusters 4, 5 and 12 (found by searching leaves) died before or soon after hatching, so the Anthanassa species were unknown for these clusters. Clusters 4 and 12 were found on Hypoestes, and cluster 5 was found on J. valerii. These data were not included in Table 2.

At least one A. ardys survived on each of the acanths used in this study, except for Hypoestes (Table 2). Although we found two egg clusters on Hypoestes (the two “unknown” egg masses were found on Hypoestes), no larvae survived past the first instar on this plant. The larvae did not appear to feed on Hypoestes. Although they were provided with both young and mature leaves, all larvae on Hypoestes died within 4-7 days. Survival rates varied among different acanth species: larvae seemed to be most successful on Blechum, Dicliptera, Justicia spp., and Pseuderanthemum (Table 2).

At least one A. tulcis survived on eight of the ten acanths used. No larvae survived past the first instar on Razisea spicata or Hypoestes, even though we observed oviposition and collected one A. tulcis egg cluster on Hypoestes. All larvae on these plants died within four days, not developing beyond the first or second instar. Again, survival rates varied among the larvae on different plants.

The one larva of Anartia fatima we attempted to rear on Hypoestes did not survive, but the one fed Dicliptera reached adulthood. We found no A. fatima eggs or larvae on Dicliptera. However, it is listed as a host by DeVries (1987).

When larvae of Siproeta epaphus were given a choice between Hypoestes, Dicliptera and Blechum, they fed only on Blechum.

Discussion

Oviposition Behaviors

Orientation toward and selection of potential oviposition sites by lepidopteran females is mediated by a combination of chemical and visual stimuli (Chew & Robbins 1984, Papaj 1986, Renwick & Chew 1994). Visual stimuli include leaf shape and color. Searching Anthanassa females landed on both acanths and non-acanths. Most often, the plants on which they landed were similar to the acanths in leaf shape (ovate), hinting that prior to landing on plants, these butterflies used visual cues to search for hosts.

The representatives of three genera of butterflies observed (Anthanassa, Siproeta epaphus, and Anartia fatima) all exhibited different oviposition behaviors. Anthanassa spp. invested more time for each oviposition event. Eggs were laid in clusters, which may serve as protection from predators or parasites (Haber 1978, Schmidt & Smith 1985), perhaps by reducing the surface area exposed to ovipositing parasitoids. Some lepidopteran larvae that feed in groups also have been shown to stimulate each other to feed (Chew & Robbins 1984). Females remained still during egg deposition, so unless females are tracked to the oviposition site, both laying females and egg clusters could be difficult for visually-oriented predators to find.

Siproeta laid eggs singly, relatively rapidly, and “on the move.” In some instances, the host plants appeared to be too small or structurally weak to support the weight of the butterfly for more than a second or two. Eggs were laid where they were difficult to see in curled leaves, at leaf nodes, and between flower bracts. This may help hide them from visual predators. Also, many Blechum plants were only 5-10 cm tall, and therefore possibly too small to support more than one or two larvae.

Anartia fatima exhibited two different behaviors—one very similar to that of Siproeta. In two locations in Monteverde, females laid eggs near, but not on their host plants. Eggs were instead laid on small Apiaceae or Rubiaceae growing in close proximity to Blechum. This type of oviposition behavior was documented in Anartia (Silberglied 1983 and included references), in Ithomiinae (Haber 1978, this study), in satyrid butterflies (Singer 1984) and in Papilionidae (Young 1979). Young (1979) and Singer (1984) speculate that this strategy protects eggs from predators or parasites that search for eggs using visual or phytochemical cues of the host plants. There is evidence that some parasitoids may use phytochemicals to locate their hosts (Hendry et al. 1976).

Interspecific constraints to larval host selection in Anthanassa spp.

Our observations indicated no apparent partitioning of resources or microhabitats (shade or sun) between the Anthanassa ardys and A. tulcis. Both species oviposited on Hypoestes plants in fairly close proximity and under apparently similar levels of light. Also, both species oviposited on Dicliptera on the same hectare of land. This could either mean that there are some unobserved differences in the microhabitats used by A. ardys and A. tulcis, or that the abundance of host plants was not limiting enough to lead to partitioning of resources.

Neither species was observed to oviposit on Blechum growing in the same habitat. S. epaphus and A. fatima laid eggs only on (or near) this species: they were not observed to oviposit on Dicliptera, Hypoestes, Justicia valerii, or Pseuderanthemum growing nearby. Resource partitioning between related species using different larval hosts has been documented for papilionids by Emmel & Emmel (1969) and Shapiro and Carde (1970). However, our data are insufficient to suggest competition or resource partitioning between Anthanassa and these other two species. Data on host use by Anthanassa species in other habitats and in other parts of their range are needed before reaching further conclusions.

Physiological and environmental constraints to oviposition site selection

Both Anthanassa species oviposited on plants along forest edges and in clearings (many acanth species grew in partial shade). Also, Anthanassa ardys and A. tulcis basked for 2-3 hours before beginning host plant searches. The shaded forest may be too cold for them. The possibility of thermoregulatory constraints to butterfly-hostplant use is discussed in Courtney (1982) and Renwick & Chew (1994). Since Anthanassa rarely ventured into the cooler forest, they would be less likely to encounter the shade-adapted species e.g., Razisea spicata, Habracanthus blephororhachis, Justicia costaricana, and J. oerstedii. This may explain why no oviposition was observed on these shade tolerant species. Blechum was most often observed growing in small patches or patches of very small plants in pastures or banana groves. It is possible that these plants were not sufficiently abundant or that resource partitioning (mentioned above) may be occurring.

Williams et al. (1996) found that females of Phyciodes tharos (Nymphalidae) oviposited on a range of possible hosts, while Chlosyne harrisii females oviposited on one host only, even though their larvae fed successfully on a range of hosts in the lab. Phytochemical as well as environmental (e.g. temperature) cues may be involved in the restricted oviposition range of Chlosyne harrisii. Phytochemistry plays a role in determining intraspecific differences in host plant ranges for some pierids (Huang and Renwick 1993).

Why did A. ardys and A. tulcis oviposit on Hypoestes phyllostachya?

Oviposition “mistakes” are documented in many lepidopteran species, such as Anartia and various ithomiids (Haber 1978 and pers. obs.). These “mistakes” sometimes involve females ovipositing on unrelated plants growing in close proximity to established hosts (as with Anartia fatima, discussed above). This has been documented by Courtney (1982), Neck (1973), and Singer (1984). Other “mistakes” involve females ovipositing on related plants that do not support larval development. These plants are often introduced species (Bowden 1971, Chew 1977 and 1981, Sevastopulo 1964, Straatman 1962).

Hypoestes phyllostachya was introduced into the Monteverde area as an ornamental plant between 1958 and 1966. The time during which Anthanassa spp. have been exposed to this plant species in Monteverde (a maximum of 40 years) may have been too short to allow Anthanassa larvae to adapt to Hypoestes either by rejecting this plant as an oviposition site or by using it as a viable host. Though the visual and/or phytochemical cues are present to stimulate oviposition by Anthanassa ardys and A. tulcis, phytochemicals that stimulate feeding behavior in larvae may be absent, or phytochemicals that are present in Hypoestes but are absent in the native acanths may be toxic to the larvae. It appeared that the Anthanassa larvae did not feed on either young or mature leaves of the Hypoestes plants to which they were exposed. In addition, early instar larvae of Anthanassa spp. appear to be rather sessile, so it seems unlikely that larvae might crawl from Hypoestes to a normal host. Thus, it seems unlikely that this behavior in Anthanassa is similar to the non-host ovipositions of Anartia fatima.

Survival rates of Anthanassa spp. on various Acanthaceae

Survival rates of Anthanassa varied among the acanth species studied, but due to the small number of larvae reared on each plant, these findings do little more than suggest possible trends for further study. The larvae reared were most successful on Blechum, Dicliptera, Justicia spp., and Pseuderanthemum (Table 2).

Conclusions

Various factors may influence the oviposition site choices of Anthanassa butterflies. Anthanassa spp. may overlook (or simply not encounter) some potential host plant species on which they can survive (e.g., in the forest, shade or low temperatures may inhibit oviposition behavior). Alternatively, they oviposit on at least one species of Acanthaceae that is unsuitable for larval development (e.g., Hypoestes). However, more study is needed to determine whether acanth host plant resources are partitioned by the butterflies at Monteverde. It would also be useful to determine exactly what visual and/or phytochemical cues stimulate Anthanassa spp. to select oviposition sites and initiate oviposition, what chemical differences exist between Hypoestes and the other acanths at Monteverde, and what effects this plant might have (if any) on populations of Anthanassa at this site.

Acknowledgments

We extend our thanks to Robert Stevenson for encouragement and support, to John Campbell and Eston Rockwell for permission to study on their farms, to the Villegas family for providing space to rear butterflies, and to Dorothy Rockwell, Sue Trosell and Lucky Guindon for information on the introduction of Hypoestes to Monteverde. We also thank Dr. Thomas C. Emmel, Stuart Feldman, Sara Feldman, and the anonymous reviewers for critiquing a draft of this manuscript and Dr. Carmine Lanciani for help with the tables. We thank Hugo Kons Jr. for his helpful comments on a later draft. Gracias a Ester Quintana por su ayuda con la revisión del resumen.

References Cited

Bolaños, R. A., and V. Watson 1993. Mapa Ecológico de Costa Rica. Centro Científico Tropical, San José, Costa Rica.

Bowden, S. R. 1971. American white butterflies (Pieridae) and English food-plants. J. Lepid. Soc. 25: 6-12.

Chew, F. S. 1977. Coevolution of pierid butterflies and their cruciferous foodplants. II. The distribution of eggs on potential foodplants. Evolution 31: 568-579.

Chew, F. S. 1981. Coexistence and local extinction in two pierid butterflies. The Amer. Nat. 118: 655-672.

Chew, F. S., and R. K. Robbins. 1984. Egg laying in butterflies. pp. 65-79 in Vane-Wright, R. I., and P. R. Ackery, eds. The Biology of Butterflies. Academic Press, London.

Courtney, S. P. 1982. Coevolution of pierid butterflies and their cruciferous foodplants IV. Crucifer apparency and Anthocharis cardamines (L.) oviposition. Oecologia 52: 258-265.

De Vries, P. J. 1987. The Butterflies of Costa Rica and their Natural History. Princeton University Press, Princeton.

Emmel, T. C., and J. F. Emmel. 1969. Selection and hostplant overlap in two desert Papilio butterflies. Ecology 50: 158-159.

Haber, W. A. 1978. Evolutionary ecology of tropical mimetic butterflies (Lepidoptera: Ithomiinae). Doctoral dissertation, Univ. of Minnesota.

Hendry, L. B., J. K. Wichmann, D. M. Hindenlang, K. M. Weaver, and S. H. Korzeniowski. 1976. Plants—the origin of kairomones utilized by parasitoids of phytophagous insects? J. Chem. Ecol. 2(3): 271-283.

Huang, X., and J. A. A. Renwick. 1993. Differential selection of host plants by two Pieris species: the role of oviposition stimulants and deterrents. Entomol. Exp. Appl. 68: 59-69.

Neck, R. W. 1973. Foodplant ecology of the butterfly Chlosyne lacinia (Geyer) (Nymphalidae). I. larval foodplants. J. Lepid. Soc. 27: 22-33.

Papaj, D. R. 1986. An oviposition “mistake” by Battus philenor L. (Papilionidae). J. Lepid. Soc. 40: 348-349.

Renwick, J. A. A., and F. S. Chew. 1994. Oviposition behavior in Lepidoptera. Ann. Rev. Entomol. 39: 377-400.

Schmidt, J. M., and J. J. B. Smith. 1985. The mechanism by which the parasitoid wasp Trichogramma minutum responds to host clusters. Entomol. exp. appl. 39: 287-294.

Scott, J. A. 1986. The Butterflies of North America: a Natural History and Field Guide. Stanford University Press. Stanford.

Sevastopulo, D. G. 1964. Lepidoptera ovipositing on plants toxic to larvae. J. Lepid. Soc. 18: 104.

Shapiro, A. M., and R. T. Carde. 1970. Habitat selection and competition among sibling species of satyrid butterflies. Evolution 24: 48-54.

Silberglied, R. 1983. Anartia fatima. pp. 682-683 in Janzen, D. H. ed. Costa Rican Natural History. University of Chicago Press, Chicago.

Singer, M. C. 1984. Butterfly-hostplant relationships: host quality, adult choice, and larval success. pp. 81-88 in Vane-Wright, R. I., and P. R. Ackery, eds. The Biology of Butterflies. Academic Press, London.

Straatman, R. 1962. Notes on certain Lepidoptera ovipositing on plants which are toxic to their larvae. J. Lepid. Soc. 16: 99-103.

Stiles, F. G. And A. F. Skutch. 1989. A Guide to the Birds of Costa Rica. Cornell University Press. Ithaca.

Williams, E. H., E. J. Stankus, and C. A. Beach. 1996. Oviposition is more restricted than feeding in use of Aster hostplants by two nymphalid butterflies. Bull. of the E. S. A. 77(3) supplement part 2: 483.

Young, A. M. 1979. Oviposition of the butterfly Battus belus varus (Papilionidae). J. Lepid. Soc. 33(1): 56-57.

 Table 2. Results from rearing Anthanassa ardys and A. tulcis on various acanth species.

Butterfly Species
Cluster #
Oviposition Site Plant
Reared on
Number of Eggs¹
Number of Larvae
Number of Adults
A. ardys²
1,2,3,6
B. glandulosa⁴
B. pyramidatum
146 total
(48, 42, 1, 55)
10
7
" "
" " " "
" "
B. glandulosa
13
6
" "
" " " "
" "
D. unguiculata
19
6
" "
" " " "
" "
H. blepharorhachis
10
3
" "
" " " "
" "
H. phyllostachya
20
0
" "
" " " "
" "
J. costaricana
14
6
" "
" " " "
" "
J. oerstedii
20
4
" "
" " " "
" "
J. valerii
16
3
" "
" " " "
" "
P. cuspidatum
15
7
" "
" " " "
" "
R. spicata
10
1
A. ardys
8,9
H. phyllostachya
H. phyllostachya
67, 35
no data⁵
0
A. ardys
10
D. unguiculata
D. unguiculata
42
no data
22
A. ardys
11
J. valerii
J. valerii
55
no data
18
A. ardys
14
D. unguiculata
D. unguiculata
28
defunct³
A. tulcis
7
H. phyllostachya
B. pyramidatum
73 total
7
1
" "
"
" "
B. glandulosa
7
2
" "
"
" "
D. unguiculata
7
2
" "
"
" "
H. blepharorhachis
4
1
A. tulcis
"
H. phyllostachya
H. phyllostachya
7
0
" "
"
" "
J. costaricana
7
1
" "
"
" "
J. oerstedii
8
8
" "
"
" "
J. valerii
7
2
" "
"
" "
P. cuspidatum
5
2
" "
"
" "
R. spicata
5
0
A. tulcis
13
D. unguiculata
D. unguiculata
96 total
41
18
"
" "
H. phyllostachya
20
0