The egg, larva and pupa of Teriocolias zelia andina Forbes, 1928 (Lepidoptera, Pieridae, Coliadinae) are described and illustrated for the first time based on specimens collected on Senna birostris var. arequipensis (Fabaceae) on the western slopes of the Andes of northern Chile. The morphology of the egg and first instar enables separating T. z. andina from species of its sister genus Leucidia Doubleday, 1847. The larva of T. z. andina passes through five instars, which can be accurately identified by head width.
Teriocolias Röber, 1909 (Lepidoptera: Pieridae: Coliadinae) is a monotypic genus of the Neotropical Coliadinae whose only species, T. zelia (Lucas, 1852), currently includes four subspecies which are mostly associated with the Andes of Argentina, Bolivia, Chile and Peru (Lamas 2004). Molecular phylogenetic studies have confirmed a close evolutionary relationship of the genus with the also Neotropical Leucidia Doubleday, 1847 (Braby et al. 2006, Wahlberg et al. 2014), as suggested earlier by Klots (1928) based on morphological characters of the adult stage.
Teriocolias zelia andina Forbes, 1928 (Figs. 1–6) was originally described from Peru (Forbes 1928). It is the only subspecies of T. zelia recorded in the Andes of northern Chile (Peña & Ugarte 1996). Although it is one of the most characteristic butterflies of the prepuna belt of the Parinacota Province (Benyamini 1995), its biology remains insufficiently known. Based on the few biological data available, T. z. andina appears to be a highly host-specialized (monophagous) butterfly, at least at the local level, because its eggs have been found only on the native shrub Senna birostris var. arequipensis (Vogel) H. S. Irwin et Barneby (Fabaceae) despite intensive field surveys, while the recently eclosed first instar is unable to feed on leaves of other native or exotic Fabaceae plants growing in this arid landscape (Vargas 2012). Furthermore, although sometimes eggs are laid on mature leaves of the host plant by the females, egg-laying is performed preferentially on new leaves (Vargas & Benítez 2013). It has been shown that the egg phenology of T. z. andina is mostly associated with the availability of plant substrate adequate for egg-laying and subsequent larval feeding (Vargas & Benítez 2013).
The importance of morphological studies of immature stages for the understanding of the systematics and evolution of Lepidoptera has been widely recognized (Scoble 1995). Furthermore, the close relationship between T. z. andina and S. b. arequipensis suggests that this butterfly-plant system could be used as a model for insect phenological studies in the arid environments of the Andes. However, the immature stages of the butterfly remain undescribed, hindering their identification in the field. Accordingly, the morphology of the egg, larva and pupa of T. z. andina are described and illustrated for the first time.
Material and Methods
Eggs of T. z. andina were searched for on plants of S. b. arequipensis near Socoroma village (18°16′S, 69°35′W), at about 3,300 m elevation on the western slopes of the Andes of the Parinacota Province of northern Chile between April 2011 and September 2015. The site is characterized by a tropical xeric bioclimate with about 9° C annual mean temperature and about 160 mm annual precipitation mostly concentrated between December and February (Luebert & Pliscoff 2006). The vegetation cover is seasonal, reaching higher levels after rains, during April–May (Muñoz & Bonacic 2006).
Leaflets with eggs were collected and brought to the laboratory in individual plastic vials, where they were maintained until larvae emerged. Fresh leaves of the host plant were added daily ad libitum until the last instar stopped feeding to start to prepare for pupation. Ten immature specimens of each instar and stage were stored in ethanol 70% and were used to perform metric measurements with the aid of a graduated ocular mounted on a stereomicroscope. Vouchers were deposited in the “Colección Entomológica de la Universidad de Tarapacá” (IDEA), Arica, Chile. The remaining individuals were reared through the life cycle to obtain adults to confirm the taxonomic identification. The adults obtained were mounted following standard procedures and were deposited in IDEA.
Results
Life history. The larva passes through five instars, which are all folivorous. The first instar makes a hole sub-apically to exit the chorion and consumes it variably, after which it begins to consume the leaflet. The leaflet consumption begins thereafter. The first and second instars scrape the leaflet, leaving the cuticle of the opposite side intact, similar to leaf skeletonizer larvae. In contrast, leaflets are completely consumed by the third, fourth and fifth instars. The fifth instar stops eating 1–2 days before pupation, moving toward shoots or raquis to pupate. The pupa is secured to the substrate by silk threads spun on the cremaster, and by a silk girdle surrounding the anterior abdominal segments and the wings.
Egg (Fig. 7, 8). Fusiform, upright, white immediately after deposition, subsequently orangeyellow. Chorion weakly striated longitudinally by about 40 ridges, most of which extend the length of the chorion, some are interrupted close the apex; an undetermined number of poorly differentiated transverse ridges; translucent, the larva may be seeing before eclosion. Height: 1.34 mm (1.30–1.40 mm); width: 0.42 mm. (0.38–0.46 mm); n = 10. Duration: 5–6 days (n = 10).
First instar (Fig. 9). Head black, thorax and abdomen yellow immediately after eclosion, greenish yellow after meal; legs, prothoracic shield, anal shield, pinnacles and setae black. Maximum length: 2.5–3.0 mm (n = 10). Duration: 6–7 days (n = 10).
Second instar (Fig. 10). Head yellowish brown; thorax and abdomen pale green, with many translucent secondary setae. Maximum length: 4.5–4.8 mm (n = 10). Duration: 4–5 days (n = 10).
Third instar. Color similar to second instar. Maximum length: 6.0–6.7 mm (n = 10). Duration: 4–5 days (n = 10).
Fourth instar. Color similar to fifth instar. Maximum length: 9.5–11.0 mm (n = 10). Duration: 4–7 days (n = 10).
Fifth instar (Fig. 11). Head, thorax and abdomen light green; small light gray dots on thorax and abdomen which are sometimes absent; a white longitudinal stripe from prothorax to A10. Maximum length: 18.0–19.0 mm (n = 10). Duration: 7–9 days (n = 10).
Instar identification. As described above, the only instar clearly different from any other is the first, based on body coloration and shape. Using the same attributes the second instar may be easily confused with the third, while the fourth instar may be confused with the fifth. However, the cephalic widths of successive instars are clearly different with no overlaps among them. Thus, this measurement may be used successfully for instar identification in T. z. andina (Table 1).
Table 1.
Mean and standard deviation (SD), interval of variation (IV) and growth rates (GR) of the head capsule width in larval instars of Teriocolias zelia andina Forbes reared on Senna birostris var. arequipensis (Fabaceae).
Pupa (Fig. 12, 13). Translucent integument; cream white initially; wing pattern of the adult may be easily observed through the integument before eclosion. Head with a conspicuous anterior projection; wings broadly projected ventrally and flattened laterally. Length: 14.0–15.0 mm (n = 10). Duration: 10–15 days (n = 10).
Discussion
Knowledge of morphology and life history of immature stages of Lepidoptera is useful in studies of systematics and evolution (Freitas 2006, Kaminski & Freitas 2008, Dias et al. 2015, Salik et al. 2015, Neves & Paluch 2016) being also a valuable tool in ecological studies requiring field identification of immature stages (Pessoa-Queiroz & Diniz 2007, Nelson 2015; Silva et al. 2016). However, although the immature stages of several Neotropical Pieridae have been described (Shapiro 1979, 1989, 1991; Aiello 1980; Braby & Nishida 2007, 2011; Kaminski et al. 2012, Hernández-Mejía et al. 2014, 2015) these remain unknown for many genera and species.
This is the first description of the immature stages of T. z. andina. The morphological pattern generally fits that described for the sister genus Leucidia (Freitas 2008); however, some differences were detected. Freitas (2008) described the egg of Leucidia as four times longer than wide, with 16–18 weakly marked longitudinal ridges and 30–35 transverse ridges. Contrastingly, the egg of T. z. andina is 3.19 times longer than wide, with the chorion finely sculptured by a large number of longitudinal ridges and an undetermined number of poorly differentiated transverse ridges. Interestingly, the morphology of the chorion of T. z. andina appears to be different to that of other Coliadinae genera already described in detail by Hernández-Mejía et al. (2014).
The first instar of Leucidia and T. z. andina also can be differentiated: this is completely green or pale green in Leucidia (Freitas 2008), contrasting with the black coloration of the head, pinnacles and setae in Teriocolias. Although no obvious morphological differences were detected for the subsequent instars, an important difference was found in development, since five larval instars were observed in T. z. andina, while the larvae of Leucidia pass through only four instars (Freitas 2008). As mentioned above, the instars of T. z. andina can be accurately separated by the measurement of the cephalic width. Accordingly, further ecological studies with this species can be complemented with detailed age characterizations of samples.
The results obtained in this study for the egg and first instar suggest that the morphological characters of immature stages may be helpful in further comparative studies of Coliadinae. However, it is evident that more detailed analyses, including scanning electron microscopy (SEM), are needed to know better the external morphology of T. z. andina, because SEM is useful either to find subtle diagnostic characters among species with highly conserved morphology and to find characters defining close genera of butterflies (Duarte & Robbins 2009; Vargas et al. 2014). Studies should be expanded to the other subspecies of T. zelia and also to additional genera of Coliadinae close to Terocolias. Furthermore, the incorporation of additional genera of Coliadinae in morphological studies of immature stages will be highly valuable to assess the character evolution in this butterfly group in the light of current phylogentic hypotheses (Braby et al. 2006, Wahlberg et al. 2014).
Acknowledgments
I thank two anonymous reviewers for valuable comments and suggestions on a preliminary version, Marcelo Vargas-Ortiz for editing the figures, Lafayette Eaton for checking the English, and Darli Massardo and André Victor Lucci Freitas for providing literature. Financial support was obtained from Project DIEXA-UTA 9710-10 from the Universidad de Tarapacá.