First Description of Female and Larva of Phyllocycla basidenta Dunkle, 1987 (Anisoptera: Gomphidae), and Ecological Niche Modeling of Phyllocycla species in Northwestern Argentina

© 2025 Academia Sinica, Taiwan Open Access First Description of Female and Larva of Phyllocycla basidenta Dunkle, 1987 (Anisoptera: Gomphidae), and Ecological Niche Modeling of Phyllocycla species in Northwestern Argentina Carlos Molineri1,* , José S. Rodríguez1, and Carolina Nieto1 1Instituto de Biodiversidad Neotropical (IBN), Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET), Facultad de Ciencias Naturales e Instituto Miguel Lillo, Universidad Nacional de Tucumán, Yerba Buena, Tucumán, Argentina. *Correspondence: E-mail: [email protected] (Molineri) E-mail: [email protected] (Rodriguez); [email protected] (Nieto) Received 8 November 2024 / Accepted 6 April 2025 / Published 25 July 2025 Communicated by Jen-Pan Huang We present the first description of female adult and larva of Phyllocycla basidenta Dunkle 1987 (Anisoptera: Gomphidae), collected in northwestern Argentina. The female and larva are described in detail, with diagnostic features that differentiate P. basidenta from other species like P. argentina and P. viridipleuris. Larvae of P. basidenta inhabit sandy-bottomed rivers alongside P. argentina. Additionally, we predict the potential distribution of both species using Ecological Niche Models under current and future climate scenarios. The models suggest that suitable habitats may shift to higher altitudes under moderate climate change, but significant losses are expected under severe warming scenarios. Climate change may threaten Phyllocycla basidenta and P. argentina as its suitable habitats could shift to higher elevations. Conservation strategies should prioritize areas projected to become suitable under future conditions. Key words: Odonata, Club tail dragonfly, Phyllocycla argentina, P. viridipleuris, Potential distribution, Ecological niche models Citation: Molineri C, Rodríguez JS, Nieto C. 2025. First description of female and larva of Phyllocycla basidenta Dunkle, 1987 (Anisoptera: Gomphidae), and ecological niche modeling of Phyllocycla species in northwestern Argentina. Zool Stud 64:19. doi:10.6620/ZS.2025.64-19. BACKGROUND Gomphids, or clubtail dragonflies, are powerful fliers often seen patrolling rivers. As adults, they are challenging to capture with aerial nets, leading to the development of alternative techniques, such as modified Malaise traps, to address this limitation (Almeida et al. 2013). The distribution range is a crucial aspect in assessing the conservation status of a species. However, for dragonflies known only from their adult stage, this range is sometimes limited to the type locality and a few scattered locations (Collins and McIntyre 2015). In contrast, their larval stage is much easier to collect, highlighting the need for increased field and laboratory efforts to associate both life stages. Phyllocycla Calvert, 1948 exemplify this situation: while the large adults are difficult to collect, larvae are commonly encountered in ecological and biodiversity studies (Araújo et al. 2020). Six species of Phyllocycla are known to occur in Argentina (Lozano et al. 2020): Phyllocycla argentina (Hagen in Selys, 1878), which is widely distributed in the Central, NE, and NW regions; P. basidenta Dunkle, 1987, found in Salta and Jujuy in the NW; P. foliata Belle, 1988, from Misiones province in the NE; P. propinqua Belle, 1972, from Entre Ríos and Misiones in the NE; P. vesta Belle, 1972, from Buenos Aires; and P. viridipleuris (Calvert, 1909), which is widely distributed in the Central and NE regions, with an unconfirmed record from Salta in the NW. The larvae of only two of these species have Zoological Studies 64:19 (2025) doi:10.6620/ZS.2025.64-19 1 © 2025 Academia Sinica, Taiwan been described: P. argentina by Rodrigues Capítulo (1983), and P. viridipleuris by Belle (1992). The larval stage of P. propinqua is reported in the literature (e.g., Belle 1992; Lozano et al. 2020), but these reports reference Needham’s (1940) description under the name P. viridipleuris (a misidentification of P. propinqua sensu Belle 1992). Needham’s description is limited to size and a few brief, uninformative sentences, which are insufficient for species characterization. Additionally, the two exuviae reported were not reared, and their association with adults was based solely on the collection site. Thus, there is no justification for considering this species’ larval stage as known. To aid in evaluating the conservation status of Phyllocycla species and knowing the variables important in their spatial occupancy, we aimed to identify all larvae collected in our study region (NW Argentina) to the species level. With this information, combined with known adult ranges, we assessed potential and future changes in species distribution. Ecological niche models (ENMs) have emerged as vital tools for understanding and predicting species distributions (Hassall 2012). By correlating environmental variables with species occurrence data, ENMs can identify areas of suitable habitat under present and future climatic conditions. This approach is particularly important for appreciating potential distribution shifts due to climate change, aiding in the conservation and management of species likely to be impacted by environmental changes (NavaBolaños et al. 2021). Phyllocycla basidenta was originally described from a male adult from Bolivia, with von Ellenrieder and Garrison (2007) later reporting its presence in neighboring areas of NW Argentina. Until now, the female adults and larvae of this species were unknown. We have reared numerous specimens, including both males and females, of P. argentina and P. basidenta in various localities of NW Argentina. Based on these reared specimens, we were able to associate the female adult and larva of P. basidenta with the male adult. Here, we present a comparative description of the previously unknown female adult and larva of this species, along with additional observations and measurements of P. argentina. We additionally provide a key to the larvae of the three species known from Argentina, as well as the potential distribution and possible future scenarios for both species. MATERIALS AND METHODS Studied material Adults, larvae and larval exuvia of reared specimen of Phyllocycla argentina, Phyllocycla basidenta, and Phyllocycla viridipleuris were studied (see details in Supplementary materials). Taxonomic work. Larvae were collected using kick nets (1 m width, 0.5 m height, 1 mm pore size) on submerged sand/mud patches in streams and rivers. The larvae of the last instars were transported on wet pieces of filter paper in plastic petri dishes. In the lab, we prepared an aquarium (0.5 m length, 0.4 m width, 0.2 m height) with a sandy substrate and water from the stream. A small aquarium pump ensured water movement and oxygenation. Larvae from the same locality were released into the aquarium until emergence. If necessary, the larvae were fed with benthic macroinvertebrates, mainly Chironomidae and small Ephemeroptera. We placed twigs and macrophytes to facilitate larval emergence. The aquarium was covered with a thin mesh. Each day, we checked for teneral adults and molts. Emerged adults were kept alive for 1–2 days and then preserved in 80% alcohol along with the corresponding exuviae. All material is housed in the Odonatological collection of the Instituto de Biodiversidad Neotropical (IBN, CONICET-UNT, Tucumán, Argentina). For each species, the coordinates are given in decimal format, and dates are recorded as day/month/year. The mandibular formula follows the proposal of Watson (1956). Photographs were taken with a Zeiss Axiocam ICc5 mounted on a Stemi 508 stereomicroscope. Drawings were made using an Olympus BX51 microscope with a camera lucida. Potential and future distribution To minimize the spatial autocorrelation between occurrences, all records in a 5 km radius were filtered and reduced to one (Cobos et al. 2019). 15 occurrence records for Phyllocycla argentina and 20 for Phyllocycla basidenta were compiled. The databases were split into 50% for training and 50% for testing the analyses. The geographical distributions of both species were estimated through correlative ecological niche modeling (ENM), ENMs establish a connection between species presence records and the environmental variables of specific locations, thereby providing an estimation of environmental suitability (Soberón and Peterson 2005). For each species, the same calibration area was designed along the Southern Yungas, extending from 150 m a.s.l. to 4810 m.a.s.l. The nineteen bioclimatic variables from WorldClim database (Hijmans et al. 2005) were utilized, with a spatial resolution of 30 seconds. The environmental data layers were clipped to the calibration area defined. A selection of variables was performed with correlation analysis to eliminate highly correlated variables (r ≥ 0.80) using the “ntbox” page 2 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan R package (Osorio-Olvera et al. 2020). For P. argentina, eight variables were selected: BIO1 = Annual Mean Temperature, BIO2 = Mean Diurnal Range (Mean of monthly (max temp - min temp)), BIO3 = Isothermality (BIO2/BIO7), BIO4 = Temperature Seasonality (standard deviation ×100), BIO9 = Mean Temperature of Driest Quarter, BIO11 = Mean Temperature of Coldest Quarter, BIO12 = Annual Precipitation, BIO13 = Precipitation of Wettest Month. For P. basidenta, twelve variables were selected: the same eight ones mentioned above and BIO7 = Temperature Annual Range (BIO5-BIO6), BIO14 = Precipitation of Driest Month, BIO17 = Precipitation of Driest Quarter, and BIO18 = Precipitation of Warmest Quarter. The calibration and selection of the best ENMs were made using a maximum entropy method with Maxent v3.4.4. K (Phillips et al. 2006) through the R package “kuenm” (Cobos et al. 2019). Candidate models were obtained by combining four values of regularization multiplier (0.1, 0.5, 1, 2), and all possible combinations of four feature classes (linear = l, quadratic = q, product = p, hinge = h). The candidate model selection and performance were evaluated based on significance (partial ROC, with 100 iterations and 50 percent of data for bootstrapping), omission rates (E = 5%), and model complexity (AICc). The best models were selected according to the following criteria: (1) significant models with (2) omission rates ≤ 5%. Then, from among this model set, models with delta AICc values of ≤ 2 were chosen as final models. The final models were generated by performing five bootstrap replicates with logistic outputs, using the full set of occurrences and the selected parameterizations. To determine the relative contribution of each environmental variable to the final model, we used three methods: percentage contribution, permutation importance and jackknife of regularized training gain. The final models were applied to the calibration area under future climate scenarios: RCP 2.6 and RCP 8.5 for 2021–2040 and 2081–2100. We employed an ensemble approach with two separate climate models: CMCC-ESM2 and EC-Earth3-Veg. The niche distribution was projected independently for each model. We then averaged the raster outputs from both CMCC-ESM2 and EC-Earth3-Veg for each climate scenario. We used the climatic layers from the Community Climate System Model (CCSM6). Also, we performed a MESS analysis (Multivariate Environmental Similarity Surface, Elith et al. 2010) to identify extrapolation risk areas in model transfers. The MESS can identify areas with large environmental differences, which may be at risk of extrapolation. We used this information to prevent over interpretation of the model outcomes by removing the suitability found in areas of strict extrapolation (Cobos et al. 2019). To obtain binary presence-absence maps for present and future scenarios, 10th percentile threshold was applied, using the “ntbox” package (Osorio-Olvera et al. 2020). To spatially characterize the distribution of P. argentina and P. basidenta, the proportion of its distribution along different altitudes in present and future scenarios was quantified. For this, an intersection layer of altitudes was obtained from a global 1-km resolution land surface digital elevation model (DEM) derived from U.S. Geological Survey (USGS) 30 arcsecond SRTM30 gridded DEM data created from the NASA Shuttle Radar Topography Mission (SRTM). To assess the extent of protection by the existing protected areas, we overlaid the potential distribution areas predicted by our models onto the protected area maps of Argentina and Bolivia (SERNAP 2023; APN 2024; SIGA 2024). RESULTS Phyllocycla basidenta Dunkle, 1987 (Figs. 1a–d, 2a–f, 3a, 4a–f, 5a–e) Male adult (Fig. 1a–d): Total length 45.0–48.0 mm; abdomen 34.0–37.0 mm (incl. app.); hind wing 27.0–28.5 mm; costal edge of pterostigma in fore wing 3.1–3.4 mm. Our material matches original description. Wings with 1 basal subcostal cross-vein. Second primary antenodal cross-vein the 5th to 7th in FW and the 4th to 6th in HW. Pterostigma brownish. Female adult: Total length 45.5–48.0 mm; abdomen 34.0–34.5 mm (including appendages); hind wing 28.5–30.0 mm; costal edge of pterostigma in fore wing 3.5–3.8 mm. Head mostly brown dorsally, except lateral green spots in postclypeus, and posfrons almost completely green except for a blackish median band (Fig. 2a, b). Labrum, mandibles and labium green. Occipital plate brown with straight hind margin fringed with dark brown setae somewhat shorter than plate length. Occipital plate (Fig. 2a, b) stout not so elongated transversally (the anterior margin is less than 2x the length of the plate at medial line). Thorax. Pronotum brown with green areas laterally. Pterothorax: inner pale stripes on dorsum extending medially but not reaching medial carina at anterior ends (Fig. 2a, c). Coloration of rest of pterothorax and legs as in male description except that for a darker metepimeron (dark green to brownish). Wings, legs and abdomen similar to male. Wings with 1 basal subcostal cross-vein. Second primary antenodal cross-vein the 5th in FW and the 4th to 5th in HW. Pterostigma brownish covering 3 to 6 cells of adjacent wing sector. Inferior lateral margins page 3 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan of abdominal segments 8 and 9 not expanded (Fig. 2d, e). Apical rim not distinctly marked. Length (mean of 3 specimens) of abdominal segments 7 (4.4 mm), 8 (2.7 mm), 9 (1.8 mm), 10 (1.0 mm) and cerci (1.1 mm). Vulvar lamina 0.3 of total length of ninth sternum, hind margin with a deep but broad V-shaped notch about 0.6 the length of vulvar lamina, each distal lobe brownish, rounded and covered with stiff setae (Fig. 2e). Fig. 1. Male adult, Phyllocycla basidenta (male IBN-O-7). a, Head and thorax, l.v.; b, same, thorax in d.v.; c, abdominal segments 7–10, l.v.; d, abdominal segment 10 and cerci, d.v. e–h, Phyllocycla argentina (male IBN-O-2). e, Head and thorax, l.v.; f, same, thorax in d.v.; g, abdominal segments 7–10, l.v.; h, abdominal segment 10 and cerci, d.v. page 4 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan Fig. 2. Female adult. a–f, Phyllocycla basidenta. a, Head and thorax, l.v. with detail of FW pterostigma; b, detail of head in d.v.; c, thorax, d.v.; d, abdominal segments 7–10, l.v.; e, abdominal sterna 8–9 and vulvar lamina; f, cerci, v.v. g–l, Phyllocycla argentina. g, Head and thorax, l.v. and detail of FW pterostigma; h, detail of head in d.v.; i, Head and thorax in d.v.; j, abdominal segments 7–10, l.v.; k, abdominal sterna 8–9 and vulvar lamina; l, cerci, v.v. Specimens: g, j (IBN-O-25), h–i and k–l (IBN-O-22). Abbreviation: me = metepimeron; op = occipital plate. page 5 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan Larval exuviae (last instar, n = 4, Fig 3a): Measurements (in mm, mean of 4 specimens unless indicated): Total length 26.5–29.5 (range); head length 3.1; head width 3.9; abdomen length 18.3; abdominal maximum width 4.1; abdominal segment X length 4.6. Antennomers 1–4 length (mm): 0.3, 0.2, 1.2, 0.3. Prementum length 1.3x maximum width, with convex semicircular ligula, palp robust with relatively slender movable hook (Fig 4c); hook 1.1 mm in length. Mandibular formula (Fig. 4e, f): L1234 0 a(m1234)b, R1234 y a(m12)b. Maxilla with 6 apical hooks (distal one bifid, Fig 4d). Legs. Burrowing hooks small and blunt on fore and middle tibiae. Length of segments (femur/tibia/tarsus/claw) in mm: fore leg 2.1/2.7/1.2/0.4; middle leg 2.2/2.8/1.2/0.4; hind leg 3.0/3.0/2.1/0.7. Wing sheaths (4.9 mm) reaching anterior 1/3 of abdominal tergum 4. Abdomen 0.7x the length of body, maximum width of abdomen (0.2x total abdominal length) on segments 3 to 5. Tube-like segment 10, 4.6 mm in length (incl. appendages of 0.6 mm), width at middle of segment 10, 0.9 mm. Coloration of larva (similar in different stages, n = 3): Head, prothorax and legs paler than the rest of body, mesothorax and abdoment yellowish light brown. Stiff setae on body and legs yellowish translucent. Thorax with a pair of submedian triangular blackish marks on fore margin of mesonotum (Fig. 3a). Abdomen yellowish without dark pigments but a paler median Fig. 3. Larval habitus, dorsal. a, Phyllocycla basidenta; b, P. argentina. page 6 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan band is present along terga 4–9 (Fig. 3a); abdominal segment 10 slightly darker than previous segments (Fig. 3a). Observations: The larva of P. basidenta were collected in the same patch of sand with other gomphid larvae, including Progomphus kimminsis Belle, 1973, P. phyllochromus Ris, 1918 and Phyllocycla argentina. Larvae of Phyllocycla basidenta and P. argentina were found superficially burrowing on fine sediments, including mud and very fine sand, but also on medium to coarse sand. This fine granulometry is more frequent and abundant in streams and rivers from submontane areas, and in NW Argentina both species are present in Chaco and Yungas ecoregions. Female adults are Fig. 4. A–F, Phyllocycla basidenta, male exuvia. a, habitus, d.v.; b, antennae, l.v.; c, prementum, v.v.; d, maxilla, v.v.; e, left mandible v.v.; f, right mandible, v.v. Abbreviations: ps = posterolateral spine. page 7 of 15 Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan Fig. 5. a–e, Phyllocycla basidenta. a, antennae l.v.; b, prementum with detail of rigid setae on ligula; c, left mandible with detail of molar (occlusal view); d, right mandible (occlusal view); e, abdomen, d.v. f–j, P. argentina. f, antennae l.v.; g, prementum with detail of rigid setae on ligula; h, left mandible with detail of molar (occlusal view); i, right mandible (occlusal view); j, abdomen, d.v. Abbreviations: ps = posterolateral spine. page 8 of 15Zoological Studies 64:19 (2025) © 2025 Academia Sinica, Taiwan more rarely seen than males, they have been caught with aerial net while ovipositing along the shore. Males patrol along the streams and marginal areas, sometimes perching on stones or marginal vegetation. Distribution: Bolivia, Argentina (Salta, Jujuy, Tucumán new record). Diagnosis: Female adult of P. basidenta can be distinguished from others in the genus by the combination of the following characters: 1) thoracic color pattern (Fig. 2a, c), 2) absence of lateral dilatation on abdominal segments (Fig. 2d, e), 3) pterostigma brownish covering 3 to 6 cells (Fig. 2a); 4) hind margin of vulvar lamina with a V-shaped notch (Fig. 2e) 0.3 times the length of ninth sternum, and 5) apical rim of segment 10 not differentiated from the remaining of the segment (Fig. 2f). Larva can be distinguished from other known larva in the genus by the following combination: 1) prementum length 1.3 times its width, labial palp and movable hook relatively slender and not pronouncedly curved (Fig. 4c), 2) abdomen with dorsal hooks on terga 3–9 (very small, except on 3–4, Fig. 5e), 3) posterolateral spines on abdominal segments 7–9 (Fig. 5e), 4) abdominal segment 10 relatively stout, ratio length segment 10/total length abdomen = 0.25 (Fig. 3a), 5) small submedian blackish triangular marks anteriorly to wing sheaths (Fig. 3a), abdomen without dark pigments but with paler mediolongitudinal band on segments 4–9 (Fig. 3a). Phyllocycla argentina (Hagen in Selys, 1878) (Figs. 1e–h, 2g–l, 3b, 5f–j) Male adult (Figs 1e–h): Total length 50.0–55.0 mm; abdomen 38.0–42.0 mm (incl. app.); hind wing 30.0–32.0 mm; costal edge of pterostigma in fore wing 4.5–4.6 mm. Wings with 1 basal subcostal cross-vein. Second primary antenodal cross-vein the 4th to 6th in both wings. Female adult: Total length 46.2–52.0 mm; abdomen 35.0–40.0 mm (incl. app.); hind wing 29.2–33.0 mm; costal edge of pterostigma in fore wing 4.3–4.6 mm. Length (mean of 3 specimens) of abdominal segments 7 (4.3 mm), 8 (2.7 mm), 9 (1.8 mm), 10 (1.05 mm) and cerci (1.3 mm). Occipital plate (Fig. 2g–i) elongated transversally (the anterior margin is about 3x the length of the plate at medial line). Pterothorax laterally pale (Fig. 2g), metepimeron pale (yellowish to light brown). Pterostigma yellowish covering 5 to 7 cells of adjacent wing sector (Fig. 2g). Wings with 1 basal subcostal cross-vein. Second primary antenodal cross-vein the 6th or 7th in FW and the 5th to 6th in HW. Vulvar lamina 0.2 of total length of ninth sternum, hind margin with a broad U-shaped (some specimens with a V-shaped) notch about 0.5 the length of vulvar lamina, each distal lobe yellowish to light brown, rounded and covered with stiff setae (Fig. 2k). Larval exuviae (last instar, Figs 5f–j): Measurements (mm, n = 3): Total length 31.3–33.5; head length 3.3; head width 4.1; abdomen length 22.7; abdominal maximum width 4.5 (located on segments 3-4); abdominal segment X length 6.1. Mandibular formula (Fig. 5h, i): L1234 0 a(m1234)b, R1234 y a(m123)b. Coloration of larva (similar in different stages): yellowish to yellowish light brown with blackish to dark brownish submedian marks (Fig. 3b) on abdominal terga 5–9 (some specimens with marks also on 3–4). Abdominal segment 10 with brownish pigments on basal 1/3. Observations: Figures presented by Belle (1970) from a male adult from Buenos Aires province (Argentina) previously studied by Ris (1913) show a subdistal dorsal spine basally to the subdistal protuberance on cercus. In all the material we have studied except one cercus of a male from Jujuy (Rio Zora), this spine is absent, as was also reported by Rodrigues Capítulo (1983). Another difference observed on our material is that lateral flap on segment 10 is absent in our males (present and pointed in Belle, 1970). The larvae of P. argentina described by RodriguesCapitulo (1983) from Córdoba are similar to our larvae relating dorsal median tubercles of abdomen (in both they are relatively thin and better visible on terga 3–5). Nevertheless, our material shows that median dorsal hooks are present through segments 3 to 9 (on segment 2 it takes the form of a short transverse ridge covered by long setae), on segments 6–9 it is small and spinelike (so small that frequently is covered by mud adhered on surrounding setae). Also, our larvae present an additional m denticle on right mandible in relation to what this author reported. Distribution: Argentina (Salta, Jujuy, Tucumán, Santiago, Córdoba, Misiones, Corrientes, Entre Rios, Santa Fe, Buenos Aires), Brazil (Rio Grande do Sul, Parana), Uruguay. Key to larva of Phyllocycla known from Argentina (P. propinqua Belle, 1972, P. vesta Belle, 1972 and P. foliata Belle, 1988 are not known in the nymphal stage). 1. Abdomen with small lateral spines on segments 6–9 (Fig. 5j) ... .................................................................................. P. argentina - Abdomen with small lateral spines on segments 7–9 (Fig. 5e) .. ................................................................................................... 2 2(1). Tubular abdominal segment 10 relatively stout, 0.25x total length of abdomen (Fig. 3a) .................................... P. basidenta - Tubular abdominal segment 10 long and slender, 0.37x total length of abdomen (fig. 40 in Belle 1992) .......... P. viridipleuris page 9 of 15Zoological Studies 64:19 (2025)