Seasonal Patterns of Inga spp. (Fabaceae) Hostplant Use by Butterflies (Lepidoptera: Pieridae) in Costa Rican Montane Forests

Abstract

Sympatric butterflies in the same family (Pieridae) utilize the same genus of hostplant (Fabaceae), yet some are only found as caterpillars on young foliage, while others appear to eat mature foliage. Observations conducted over the course of a single year on five species of Inga (Fabaceae) in Monteverde, Costa Rica, revealed three butterflies whose caterpillars consumed their leaves. One species was more common throughout the year with caterpillars encountered on primarily mature foliage; two others were found only on species with new leaves and occurred seasonally, during months of increased new leaf production. In feeding preference tests, those species showed a marked preference for eating new leaves, whereas the more common species showed no preference. The seasonality of occurrence of the two species eating new leaves may be explained, in part, by availability of their preferred foliage. The coexistence of the three butterflies using the same host-plants may be facilitated by their leaf age preferences and differential oviposition patterns of the females.

1. Introduction

In the cloud forests of Monteverde it is not unusual to see butterflies, both in sunny gaps and clearings and in the understory of the forest. Distinctive among these are the striking yellow sulphur butterflies (Pieridae: Coliadinae), some of which have been described as “the apotheosis of butterflyness” [1], and their less showy forest relatives, the mimic-whites (Pieridae: Dismorphiinae).

The Pieridae subfamilies Dismorphiinae and Coliadinae feed predominantly on legumes (Fabaceae) and are considered ancestral to other Pieridae [2]. Phylogenetic work using larval hostplant character shows that legumes were the ancestral hostplants of the family [3].

The foliage of Inga trees (Fabaceae: Mimosoideae) serve as hostplants for numerous Lepidoptera in the Neotropics, including species of butterflies in the family Pieridae [4]. Many Inga spp. benefit from biotic defense by ants that are recruited by extrafloral nectar produced on their leaves [5,6] as well as varied chemical defenses that have promoted diversification in this genus of plants [7]. Extrafloral nectaries are more effective as ant attractants at lower elevations and may serve to attract other beneficial insects (e.g., parasitoids) at higher elevations [8,9,10].

In plants with foliar nectaries, new leaves are typically softer and less well-protected chemically than mature leaves. Perhaps for this reason, extrafloral nectar is most important in attracting beneficial insects to protect developing leaves and is more copiously secreted on young leaves than on fully expanded and hardened leaves. Some butterflies choose to lay their eggs on the newest leaves, so that their caterpillars might find soft, nutritious tissue to eat when they hatch. Others can lay their eggs on mature leaves, and their larvae spend their early instars skeletonizing leaves by eating the tissue between veins. As leaves document their history of herbivory (unless they have been eaten entirely), it is possible to trace the development of caterpillars of those butterflies by seeing the progression from skeletonizing to edge trimming to consumption of entire leaves as they move around the plant prior to pupation.

Observations of Lepidoptera on the wing reveal different seasonal patterns of occurrence among the pierid butterflies, leading to some questions:

(1)

Why is one species more common throughout the year, while the others are only around at certain times?

(2)

Is the seasonality of sympatric pierid butterflies related to whether they require new leaves for oviposition? And is this related to rainfall patterns?

The hypothesis formed from these questions is: three sympatric butterfly species utilizing the same hostplants partition the resources to facilitate their coexistence.

2. Materials and Methods

This study was conducted across an elevational range in the Cordillera de Tilarán in central Costa Rica. These sites are in the vicinity of Monteverde (1300 m–1650 m elevation), in habitats including lower montane rain forest and lower montane wet forest/rain forest [11]. Annual temperature was 18.5 deg C, with a minimum of 9 deg C and a maximum of 27 deg C measured at 1460 m (J. Campbell data cited [12]). Average annual rainfall ranged from 1715 mm to 3240 mm, with a mean of 2519 mm [12]. Rainfall is seasonal, with December–April the drier months (Figure 1f).

At each of three study sites, ten tagged saplings of each Inga species were monitored monthly for one year, from July 1979 through June 1980. The tree taxa were Inga hintonii Sandwith (I. densiflora in pre-2000 publications); I. longispica Standl.; I. mortoniana J. Leon; I. punctata Willd.; and I. sierrae Britton & Killip (I. brenesii in pre-2000 publications). All caterpillars encountered were removed and reared [8]. A total of 70,648 leaves were examined on all species combined. Each Inga species had a different number of leaves observed (I. hintonii 1911, I. longispica 846, I. mortoniana 5819, I. punctata 1441, I. sierrae 631); the total counts of caterpillars observed on each Inga sp. each month was divided by the number of leaves examined for each species. Numbers of caterpillars per 100 leaves is therefore used as a measure across all the Inga species. Phenology of leafing, flowering, and fruiting were monitored on different marked adult Inga individuals in the same areas. Only leafing data are presented here; the others are presented elsewhere [13].

Feeding preference tests were conducted with third instar larvae of each species, to compare their preference for eating young or mature foliage of each Inga species. We performed at least ten trials for each Inga/caterpillar spp. combination, using the same Inga species from which the caterpillar had been collected in the field (these additional caterpillars were collected from plants other than those designated to monitor caterpillar numbers). We placed each caterpillar in a separate plastic box with both new and older leaves of a given Inga sp. and left them to feed for 24 h. At that time, we removed the leaves and compared the amount of each type of leaf eaten. Each leaflet consumed was traced on graph paper (20 sq/in) and the number of squares consumed recorded. Boxes were cleaned, sterilized with bleach, and allowed to dry before they were re-used. We used fresh leaves for every trial, and each caterpillar was used only once. If the caterpillar did not eat anything, but appeared healthy, we omitted that trial from the data (assuming they may have been changing instars), continuing until we got a total of ten trials.

We considered the number of squares of leaf area consumed and, if there was more than twice the amount of one leaf type consumed, considered the type more consumed to be preferred in that trial. If comparable amounts were consumed, we considered there was no preference shown. We used a sign test to compare the overall significance of any difference among the ten trials, using the one-tailed test as the expectation was that new leaves would be more palatable than mature leaves.

3. Results

3.1. Seasonality of Leaf Production

In Inga hintonii (Figure 1a), new leaf production dips to 5% and less during the dry season months of January and February (Figure 1f), but the rest of the year produced new leaves over more than 10% of its canopies. New leaves are produced throughout the year on Inga punctata; every month more than 10% of the canopy is new leaves (Figure 1d). Similarly, Inga sierrae produced roughly 10% new leaves steadily every month throughout the entire year (Figure 1e).

The two Inga species with foliage flushing red show greater seasonality of leafing: Inga longispica, growing in the highest and wettest site, produced few new leaves over the study months and showed only one period of more intense leafing during the early wet season. (Figure 1b). Inga mortoniana produced new leaves at a low level during the wetter months, with greater leaf flushing in the dry season months, especially February and March (Figure 1c).

3.2. Seasonality of Caterpillars on Study Trees

The most common caterpillars encountered were those of Dismorphia crisia lubina A. Butler, 1872, a subspecies that occurs in Panama and Costa Rica. This species oviposits on mature leaves of Inga as well as new leaves, and its larvae leave a distinctive pattern in each instar (Figure 2). We found caterpillars on all five species of Inga monitored in this study (Figure 1a–e). Like many pierid butterflies, these caterpillars pupated on their hostplants, as the chrysalis is green and is camouflaged among the leaves.

Caterpillars of Lieinix nemesis G & S (Latreille, [1813]) were only found in the spring months (April and May) on the three Inga species that produce new leaves that are green in color (I. hintonii, I. punctata, and I. sierrae) (Figure 1a–c). These caterpillars were usually found on newer leaves (Figure 3), in contrast to those of Dismorphia. Interestingly, they were never encountered on the two Inga species with new leaves that are red (see below).

We usually found Phoebis hersilia (Cramer, 1779) caterpillars on the new red leaves of I. longispica and I. mortoniana (Figure 1d,e), though they were infrequently on young leaves of I. hintonii (Figure 1a). The females apparently oviposit on the new foliage, and the caterpillars are red when eating young red leaves. In later instars they eat the older leaves and are greener in color (Figure 4).

3.3. Caterpillar Feeding Preferences

Dismorphia crisia lubina, found on all five Inga species, on both old and young leaves, showed no preference (that is, they ate similar amounts of each leaf type) for either older or younger leaves of any Inga species in our trials (

Table 1. Outcome of Feeding Preference Tests: (a) Dismorphia crisia lubina caterpillar feeding preference tests with five Inga species. (b) Lieinix nemesis caterpillar feeding preference tests with three Inga species. (c) Phoebis hersilia caterpillar feeding preference tests with two Inga species.

(a) Dismorphia crisia lubina
Inga Species	Young Leaves Preferred	Old Leaves Preferred	No Preference	Sign Test Difference (2-Tailed)
I. hintonii	14	15	2	NS, p = 0.5
I. longispica	2	7	1	NS, p = 0.09
I. mortoniana	7	2	2	NS, p = 0.09
I. punctata	12	14	3	NS, p = 0.42
I. sierrae	7	3	2	NS, p = 0.17
(b) Lieinix nemesis
Inga Species	Young Leaves Preferred	Old Leaves Preferred	No Preference	Sign Test Difference (2-Tailed)
I. hintonii	9	1	0	p = 0.011
I. punctata	10	0	0	p = 0.001
I. sierrae	10	0	0	p = 0.001
(c) Phoebis hersilia
Inga Species	Young Leaves Preferred	Old Leaves Preferred	No Preference	Sign Test Difference (2-Tailed)
I. longispica	10	0	0	p = 0.001
I. mortoniana	10	0	0	p = 0.001

a). Caterpillars were numerous enough that we ran extra trials with these caterpillars on several Inga species, and in no case was there any preference displayed.

We encountered Lieinix nemesis larvae on the young foliage of three Inga species (I. hintonii, I. punctata, and I. sierrae). These caterpillars uniformly showed a preference for young leaves (Table 1b). Similarly, we only found caterpillars of Phoebis hersilia on younger leaves of I. longispica and I. mortoniana, and in every trial they preferred young leaves over mature leaves (Table 1c) of their host species.

4. Discussion

Long-term study has found that caterpillar faunal diversity has declined in protected tropical areas due to changing climate and weather anomalies [14]. Hotspots of butterfly diversity, including tropical mountains, are threatened in a warming world [15]. Oligophagous butterflies, utilizing only a few related hostplants as do these pierids, are restricted in their abundance by the presence of hostplants. In conservation of rare species, it is essential to ensure that suitable hostplants exist [16]. Long-term butterfly population persistence depends most heavily on the within-site quality of caterpillar habitat [17]. Butterfly populations have continually declined in the US, but as butterfly populations can increase rapidly given the right circumstances, species recovery from even small populations is possible [18].

For butterflies that oviposit only on young foliage of their hostplants, it is unsurprising that they are found only when their hostplants are producing new leaves. DeVries [19] reported the same caterpillar feeding preferences as we observed and found in our feeding experiments: Dismorphia larvae start out on older leaves but eat both old and new leaves; Lieinix eats young foliage; and Phoebis only eats young leaves (Marquis observations in [19]). Our simple feeding experiments demonstrated preferences of the butterflies that agree with these earlier observations by expert lepidopterists.

The color of Phoebis hersilia caterpillars on new red leaves is likely a color sequestration that promotes camouflage against visually oriented predators. This phenomenon is known for other Phoebis species eating the yellow flowers of Senna spp. in preference to the new green foliage: caterpillars match the color of the plant material they consume [20], though adults resulting from those different colored caterpillars are indistinguishable.

Research on antiherbivore defenses of Inga spp. in Panama have discovered that sympatric species differ in defensive strategies [6] and show divergent chemical and biotic defenses more than random expectation [7]. Herbivores specializing on leaves of different developmental stages, as observed in this study, is still another way that more species may be accommodated on sympatric Inga spp. host plants.

Studies have shown that patterns of butterfly species richness and abundance are influenced by both seasonality and habitat structure [21,22]. Long-term study over many years revealed that rainfall patterns drive annual cycles in species diversity and community similarity [23], and rainfall triggers new leaf production in many plant species that corresponds to increased oviposition by specialist butterflies that preferentially oviposit on newer leaves [24,25]. It is unusual that the adult stages of butterflies span multiple seasons [26]. Many butterflies are more common, or occur exclusively in warmer weather, in both temperate and tropical [27] systems. More species are active during warmer seasons, and there exists greater diversification in places with more constant temperatures and cooler maximum temperatures in the warmer season [28]. In the present study, only one of the sympatric pierids occur in multiple seasons, while the other two are highly seasonal, and seasonality is related to new leaf production, and therefore leaf age, with young leaves only sporadically available for those that prefer young leaves for oviposition and caterpillar feeding.

It has been suggested that in some tropical forests, high parasitoid pressure primarily drives caterpillar abundance, with more caterpillars present in the early dry season than throughout [29]. Our observations of Dismorphia parasitization showed this seasonal trend, but this may also have been influenced by host caterpillar density and the number we were able to rear in our studies.

The historical data presented here are from a study done over forty years ago, part of a larger study of plant/animal interactions of Monteverde Inga spp. (both pollination and herbivory). The apparent divergent preferences of the three pierid butterflies for leaves of certain Inga species, and leaves of certain ages, warranted a closer look to try to discern what made possible their coexistence in this area. There have been striking changes in climate worldwide and it may be illuminating to study how these butterflies and their hostplant interactions have changed over the decades.

5. Conclusions

By examining a subgroup of taxonomically related herbivores on a single genus of plants, we have determined that, for some species, their seasonal occurrence corresponds to new leaf production in the host-plants. The demonstrated preferences shown for young leaves of Inga spp. may help these butterflies coexist in the same habitats with related species that are able to consume mature foliage throughout the year. Diversification of the Pieridae using Fabaceae host-plants apparently involved not only radiation to utilizing other legume plant species [2] but by evolving preferences for oviposition and feeding on leaves of different stages of development.