Chemical Cues Influence Pupation Behavior of Drosophila simulans and Drosophila buzzatii in Nature and in the Laboratory

In the wild, larvae of several species of Drosophila develop in heterogeneous and rapidly changing environments sharing resources as food and space. In this scenario, sensory systems contribute to detect, localize and recognize congeners and heterospecifics, and provide information about the availability of food and chemical features of environments where animals live. We investigated the behavior of D. simulans and D. buzzatii larvae to chemicals emitted by conspecific and heterospecific larvae. Our goal was to understand the role of these substances in the selection of pupation sites in the two species that cohabit within decaying prickly pear fruits (Opuntia ficus-indica). In these breeding sites, larvae of D. simulans and D. buzzatii detect larvae of the other species changing their pupation site preferences. Larvae of the two species pupated in the part of the fruit containing no or few heterospecifics, and spent a longer time in/on spots marked by conspecifics rather than heterospecifics. In contrast, larvae of the two species reared in isolation from conspecifics pupated randomly over the substrate and spent a similar amount of time on spots marked by conspecifics and by heterospecifics. Our results indicate that early chemically-based experience with conspecific larvae is critical for the selection of the pupation sites in D. simulans and D. buzzatii, and that pupation site preferences of Drosophila larvae depend on species-specific chemical cues. These preferences can be modulate by the presence of larvae of the same or another species.


Introduction
Drosophila adult and larva behaviors in nature have been poorly investigated [1,2]. This causes a limitation in the investigation related to population genetics, evolution and neurosciences in Drosophila genus. For example, even if we know that Drosophila larvae can learn [3], the role of learning in adaptation to changing breeding sites, where these pre-adults develop, remains poorly explored. This knowledge is important to fully understand population genetics and ecology in relation to the evolution of Drosophila genus. In particular, we have very little knowledge on the involvement of learning in food preference, feeding rate, and selection of pupation sites. Since many Drosophila larvae develop within decaying fruits, which change their composition in a relatively short time, Drosophila larval behavior may show a high flexibility and plasticity [4]. On the other hand, ecological variation of the breeding sites also may affect some aspects of Drosophila adult life cycle, as the reproduction strategies [5] and selection of oviposition sites [6].
Even if the cellular and molecular genetics bases of odor perception have been intensively unraveled in Drosophila adults and larvae [7], the identity of natural chemicals surrounding the breeding sites is virtually unknown. Several studies suggested that the selection of pupation sites in species of Drosophila breeding in the same breeding sites largely depend on the discrimination of chemicals emitted by conspecifics and heterospecifics [1,8]. Here, we focused on the selection of pupation sites by larvae of Drosophila simulans (Subgenus Sophophora; melanogaster species group) and Drosophila buzzatii (Subgenus Drosophila; repleta species group) in nature and in the laboratory. In Chile, the two species together with Drosophila nigricruria and Drosophila hydei (Subgenus Drosophila; repleta species group) cohabit in decaying prickle pear fruits (Opuntia ficus-indica) and necrotic tissues of this plant and columnar cactus Echinopsis chilensis. In Argentina, D. buzzatii lives in desertic and woodland areas in association with one or more species of Opuntia. Thus, D. buzzatii is a widespread species with colonizing ability. Drosophila pupation behavior is related with habitat selection, colonization of new niches and the expansion of populations [9,10]. Specifically, we focused on the natural phenotypic variation in pupation site preferences resulting of the presence of homo-and heterospecific individuals sharing the same breeding site. For each species, we tested the effect of larval chemicals on the selection of pupation site.

Natural Variation in Pupation Site Preferences
In their natural breeding sites, D. simulans and D. buzzatii changed their pupation site preferences in the presence of the other species. In the absence of D. buzzatii, about half D. simulans larvae pupated inside the fruit (Fig. 1), and most of the other pupae were found on the land underneath the fruit (D. simulans pupae were rarely found near-under and over-the fruit skin). When mixed with D. buzzatii, most D. simulans pupae developed inside the fruits, whereas fewer were found on the land underneath the fruits or in contact with the fruit skin. Therefore, the presence of D. buzzatii changed D. simulans pupation site preference in the nature, Fig.1; x 2 = 117.16, df = 4, P,,0.01.
Reciprocally, in the fermented fruits only colonized by D. buzzatii, most pupae were found under the skin, whereas the remaining pupae were found inside the fruit and much frequently on the skin surface. In the presence of D. simulans, most D. buzzatii pupae were found -under or over-the skin. This effect was highly significant, Fig.1; x 2 = 18.29, df = 2, P,,0.01. In conclusion, the presence of heterospecifics affected the distribution of pupae in different microhabitats of the fruit.

Larval Behavior to Conspecific and Heterospecific Larval Chemical Cues
Since our observation in nature indicated that interaction between larvae of the two species can affect their pupation site, back in the laboratory, we tested the role of chemical cues potentially involved in this effect. To asses this, we measured both the larval behavior and pupation site preferences in larvae presented to a dual-choice test consisting of two filter paper targets impregnated by the food processed by larvae of each species. First, we measured the time spent by third instar D. simulans and D. buzzatii larvae on each target paper. The performance of larvae either reared with conspecifics or in isolation was compared ( Fig.2). D. simulans larvae reared with conspecifics remained much longer on the D. simulans paper than on the D. buzzatii paper (15.6 and 5.0 min, respectively; U 0.05, 5,19 = 78; P,0.05; Fig. 2a). Reciprocally, D. buzzatii larvae reared with conspecifics spent much more time on the D. buzzatii paper than on the D. simulans paper (10.3 and 3.0 min, respectively; U 0.05, 6,12 = 61; P,0.05). By contrast, larvae of the two species reared in isolation showed no marked preferences: they spent a similar amount of time on the D. simulans and D. buzzatii papers (for D. simulans larvae: 8.7 and 10.5 min, respectively; U 0.05, 4, 6 = 2.0; P.0.05; for D. buzzatii larvae: 6.7 and 7 min, respectively; U 0.05, 2, 16 = 3.6; P.0.05; Fig. 2b). In conclusion, this experiment shows that D. simulans and D. buzzatii larvae raised with conspecifics spent more time on sites impregnated with their own species-specific chemicals.  The larvae were reared with conspecifics (Fig. 2a), and in isolation from conspecifics (Fig. 2b), and tested individually. Black columns, time spent on simulans paper; white columns time spent on buzzatii paper. *Mann-Whitney U-test; P,0.05; NS = non significant differences (see text for details). doi:10.1371/journal.pone.0039393.g002

Influence of Conspecific Larval Chemical Cues on Pupation Site Preferences
We also studied the effect of conspecifics on pupation site preferences by comparing the performance of individuals raised either alone or with conspecifics (Fig. 3). D. simulans larvae raised with conspecifics pupated much more often on D. simulans target paper (35.8%, than on D. buzzatii paper, 10.2%, U 0.05, 10, 30 = 221; P,0.05). Similarly, D. buzzatii larvae raised with conspecifics pupate preferentially on the conspecific paper than on the D. simulans paper (49.0 and 12.0%, respectively; U 0.05, 10, 35 = 249; P,0.05. By contrast, larvae reared in isolation showed no pupation site preference. More precisely, a similar frequency of pupae was observed on either D. simulans or D. buzzatii papers (for D. simulans: 32.2 and 33.8%, respectively; for D. buzzatii: 42.1 and 46.3%, respectively). In conclusion, D. simulans and D. buzzatii third instar larvae raised with conspecifics distinguished simulans and buzzatii chemical cues to select pupation sites, but they did not when bred in isolation. These results suggest that larval experience with conspecifics play a critical role in pupation site preferences in both species.

Discussion
Larvae of a wide range of Drosophila species use the same decaying fruits as breeding sites [11]. This implies that they are adapted to grow and develop in the presence of other species of the genus, and this may be reflected by their responses to chemicals emitted by other larvae. Our data support this assumption; D. simulans and D. buzzatii larvae responded to chemical cues produced by individuals of the two species. As a result the larvae changed their pupation site preferences. Most importantly, larvae of each species pupated away from the pupation sites selected by larvae of the other species. These findings are supported by previous studies [1,8], suggesting that the pupation behavior shown by D. simulans and D. buzzatii is part of larval behaviors that allow the optimal exploitation of breeding site units by Drosophila species.
On the other hand, responses to chemical cues produced by D. simulans and D. buzzatii larvae introduce a spatial dimension to pupation behavior. This means that the larval selection of pupation sites may result in space competition within the breeding sites, this affecting the expansion rate and dynamic of the populations. In other words, the spatial localization of D. simulans and D. buzzatii pupae within their breeding sites may influence the ecological, genetic and evolutionary dynamics of their populations [2,12].
Our data also show that rearing conditions change the dispersal patterns of D. simulans and D. buzzatii larvae prior to their selection of pupation sites. The fact that larvae reared alone spent less time on a paper impregnated in conspecific chemical cues than larvae of the same strain reared with conspecifics (Fig. 2 a, b) suggests that early experience with conspecifics will subsequently affect the patterns of pupae. Given the heterogeneous and changing ecological conditions of decaying prickly pear fruits where larvae develop, the detection, recognition, and spatial discrimination of conspecific and heterospecific cues may help dispersing larvae to find their way within the breeding sites. On the other hand, larval behaviors are flexible, and this flexibility depends on species, rearing conditions during early development, and ecological circumstances including the presence/absence of heterospecifics (Figs. 1-3). Then, Drosophila breeding sites provide environments where recognition of chemical cues left by conspecifics maybe crucial for the selection of pupation sites. Our data also suggest that, in the wild, stimuli provided by conspecifics and hetero-specifics are processed through some type of learning process as described in D. melanogaster larvae [3,13,14]. Further work is required to discover the mechanisms and neural pathway involved in this process.
Our study shows that pupation site preferences of Drosophila larvae depend on species-specific chemical cues. These preferences can be changed by the presence of other larvae of the same or of another species. Our data suggest that the natural sites where D. simulans and D. buzzatii pupae are found reflect a careful choice of the most appropriate place to pupate. Similar studies could help to investigate the role of biotic factors in other ecologically relevant larval behaviors such as feeding rates and food preferences. Our  (Fig. 3a), and in isolation from conspecifics (Fig. 3b). The larvae were tested individually (N = 50 per species). Black columns, percentages of pupae on the simulans paper; white columns percentages of pupae on the buzzatii paper. *Mann-Whitney U-test; P,0.05; NS = non significant differences (see text for details). doi:10.1371/journal.pone.0039393.g003 study should also provide a new insight to understand how the choice of niche can be modulated through the neural perception of chemical cues affecting pupation behavior on a short term, and perhaps driving evolutionary changes on the long term. The nature of these signals is unknown, but Drosophila larvae could produce pheromones as those involved in mate recognition in the melanogaster Subgroup [9]. Clearly, future work should focus on this problem.

Field Collection and Strains
We used both cosmopolitan D. simulans (subgenus Sophophora; melanogater species group) and widespread D. buzzatii (subgenus Drosophila; repleta species group). In Chile, the two species have a strong colonizing ability. Thus, larvae, pupae and adults can be collected from decaying fruits of prickly pear Opuntia ficus-indica between Copiapó, in the northern part of Chile (latitude 25u 159), and Curicó, in the south (latitude 34u 389). Some very few adults of D. melanogaster, D. nigricuria, D. hydei and Chilean endemic D. pavani may also emerge from the fruits and necrotic prickly pear tissue. The fruits containing pupae of D. simulans and D. buzzatii were collected in Til-Til (latitude 33u 59), 50 km North-East from Santiago in April of 2008 when Chilean populations of Drosophila reach their peak of abundance [15]. Thus, we used sympatric strains of the two species. Fruits were randomly picked up. We collected 46 fermented fruits of prickly pear containing pupae of D. simulans, other 47 with pupae of D. buzzatii, and 86 fruits with pupae of both species. We counted the number of pupae on different parts of the fruits (inside the fruit tissue, over the skin, under the skin) and on the land underneath the collected fruits (Fig. 1). Species were identified on the base of the color and size characteristics of pupae [1]. For the few unidentified cases, each pupa was individually deposited in a vial containing a piece of moistened filter paper, and stored in an incubator at 24uC until emergence for species identification. D. simulans and D. buzzatii studied lines were initiated with flies emerging from fermented prickle pear fruits (details in [1]). All lines were maintained at 24uC in half pint bottles containing 50 cc of [16].

Laboratory Tests
To measure pupation site preference and larval behavior, third instar larvae were individually placed at the center of a Petri dish filled with 3% agar with a thickness of 2 cm. On both sides of the center (at about 3 cm) of the dish, we placed two round pieces of filter paper (diameter = 2 cm). Each piece of paper was impregnated during 1 h in Burdick's medium processed by larvae of either species. For larval behavior experiment, we recorded the time spent on each type of impregnated paper by individual larva (N = 50/species). All these larvae were reared together in Burdick's medium. We also measured the behavior of 50 other individuals of the same species reared in isolation from conspecifics. After larval behavior measurement, the Petri dishes containing the impregnated papers were incubated at 24uC until pupation and the number of pupae found on each type of paper was noted (pupae found under, around and under the paper were pooled).

Statistical Analysis
To compare the distributions of pupae on different parts of the fruit, we used R x C test of independence. We used the Mann-Whitney test to compare time of permanence, and pupation site preferences of the larvae reared with/out conspecifics [17].
No specific permits were required for the described field studies from any authority. We talked with the owner of the private land, ''Fundo La Capilla'', where the field studies were made, Don Juan Ignacio Herrera and Don Gonzalo Herrera, who agreed we examined, collected decaying fruits, and the fruit flies for our field and laboratory studies. The fruit units examined were fallen on the ground; they correspond to disposable fruits unfit for human consumption. The field studies did not involve endangered or protected species.