Allorhogas conostegia Marsh and Shaw, 2009
|
publication ID |
https://doi.org/ 10.1080/00222930903243996 |
|
persistent identifier |
https://treatment.plazi.org/id/03BB4A5C-F336-BE2E-FE2F-6372768B6BA7 |
|
treatment provided by |
Felipe |
|
scientific name |
Allorhogas conostegia Marsh and Shaw |
| status |
sp. nov. |
Allorhogas conostegia Marsh and Shaw , sp. nov.
( Figure 1 View Figure 1 )
Female
Body length. 2.5–3.0 mm.
Colour. Body mostly yellowish brown to honey yellow, flagellum yellowish brown to yellowish white, gradually darker apically, apical tarsomeres brown, wing veins light brown to yellowish white, stigma yellow to yellowish white.
Head. Face granular, swollen medially ( Figure 1A View Figure 1 ); malar space granular, about onethird eye height; frons granular, deeply excavated, often weakly margined near eyes; vertex granular, ocellar–ocular distance 3.5–4.0 times diameter of lateral ocellus; temple granular, in dorsal view swollen behind eye ( Figure 1B View Figure 1 ); 18–22 flagellomeres.
Mesosoma. Pronotum laterally granular dorsally, scrobiculate below, pronotal collar very short behind eye; mesoscutal lobes granular, notauli deeply scrobiculate, meeting before scutellum in triangular rugose area, median lobe with median groove which is distinct at least in triangular rugose area, often weakly indicated anteriorly ( Figure 1C View Figure 1 ); scutellum granulate, prescutellar furrow wide, with four or five cross carinae; mesopleuron weakly granular, often almost smooth above sternaulus, subalar groove rugose, sternaulus weakly scrobiculate, about two-thirds width of mesopleuron ( Figure 1D View Figure 1 ); propodeum entirely rugose, areola distinctly margined by carinae ( Figure 1C View Figure 1 ).
Wings. Forewing vein r one-half length of vein 3RSa, vein m-cu meeting vein 2RS in holotype and most paratypes, occasionally slightly distad vein 2RS; hindwing vein M + CU equal to or slightly shorter than vein 1M.
Legs. Hind coxa with distinct anteroventral basal tubercle, hind femur about four times as long as wide.
Metasoma. First tergum short and broad, apical width slightly greater than length, basal area smooth and at almost 90° to apical half, tergum longitudinally costate beyond basal area, basal area with distinct carina separating it from apical half of tergum, median area distinctly margined laterally and with weaker median longitudinal carina ( Figure 1E View Figure 1 ); second tergum longitudinally costate; third tergum longitudinally costate on basal half, smooth on apical half; remainder of terga smooth; ovipositor ( Figure 1F View Figure 1 ) about one-half length of metasoma.
Male
Essentially as in female; mesosoma and metasoma mostly brown to yellowish brown, ventrally pale yellow to yellowish white; hind femur swollen, 2.5–3.0 times as long as wide at widest point.
Holotype female
Costa Rica: San José, Zurqui de Moravia , 1600 m, 15 September 2001, ex. Conostegia xalapensis fruits, Paul Hanson coll. Deposited in ESUW.
Paratypes
Twenty females, 64 males, same data as holotype. Deposited in NMNH (four females, 16 males) , MCZ (two females, eight males) , MZCR (one female, four males) , and remainder of series in ESUW. Seven females, two males, San José, San Jeronimo de Moravia , 17 May 2004, ex. Conostegia xalapensis fruits, K. Nishida coll. Deposited in ESUW . Nineteen females, 14 males, San José, Escazu , 15 October 2005, ex. Conostegia xalapensis fruits, L. Chavarría coll. Deposited in ESUW .
Comments
Marsh (2002) presented a key to the species of Allorhogas occurring in Costa Rica in which most species had forewing vein m-cu arising basal to or in line with vein 2RS, but a few had vein m-cu arising distal to vein 2RS. This character now appears to be more variable than previously thought because both conditions occur in the type-series of A. conostegia described above. As a result, in Marsh’s key, A. conostegia will run either to A. sulcatus Marsh , or to A. zurquiensis Marsh , depending on the position of vein m-cu. Allorhogas conostegia is easily distinguished from A. sulcatus , which has a long and distinct median mesonotal sulcus and mostly black body colour, and it can be distinguished from both of these other species by its smaller body size and fewer flagellomeres. Allorhogas conostegia is 2.5–3.0 mm long with 18–22 flagellomeres. Both A. sulcatus and A. zurquiensis are larger (body length 3.5– 4.0 mm and 3.5–4.5 mm, respectively) and both have more flagellomeres (28 flagellomeres and 24–28 flagellomeres, respectively). The shape of the metasoma is more compact in A. conostegia ( Figure 1E View Figure 1 ), whereas in A. sulcatus and A. zurquiensis the metasoma is longer and more an elongated oval in shape (as in fig. 91 of Marsh 2002).
Body colour in A. conostegia appears to be rather consistent within reared series from a single location, but varies significantly among series from different locations. The series from Zurqui de Moravia has more specimens that were a darker brown, whereas the series from San Jeronimo de Moravia contained the lightest-coloured specimens. In some individuals from San Jeronimo de Moravia, the light colours of the head, antenna, legs, wing veins, sternaulus and metasomal venter are so pale that they appear almost white.
The galls
In addition to A. conostegia , seven species of Chalcidoidea ( Hymenoptera ) were reared from the fruit galls on C. xalapensis: three species of Tetrastichinae (Eulophidae) , at least one of which is Aprostocetus ; Eupelmus sp. (Eupelmidae) ; Eurytoma sp. and Sycophila sp. (Eurytomidae) ; and Torymus sp. (Torymidae) . However, none of these other gall occupants was as consistently present in all samples as was A. conostegia , which strongly suggests that the latter is the gall inducer. In none of the fruit dissections were mature larvae or pupae of A. conostegia found together with host remains, which one would expect if it were a parasitoid (although this is not always the case). To the contrary, ectoparasitic larvae were often found on A. conostegia , although these were not individually reared to determine their identities. We therefore conclude that A. conostegia is the gall inducer and that the six chalcidoids are parasitoids and/or inquilines. Total parasitism by the six chalcidoids at Zurquí was 44%, while at Escazú it was 55%.
The galls are located inside the fruits of C. xalapensis and are not visible without dissection. The exact tissue (seed versus fruit) in which the galls are located is still uncertain. Fruits contained 1–14 galls, but the average was two or three (see below). Uninfected fruits ranged from 0.17 cm in diameter when immature to 2.92 cm when mature, while galled fruits had a maximum diameter of 2.6 cm. The galls contained only one A. conostegia per gall. When the galls were larger, they occupied from 50 to 85% of the fruit, pressing against the carpels and making the latter difficult to distinguish. In 9% of all infested fruits, different galls contained different developmental stages (larva, pupa or adult) of A. conostegia . In the carpel where the gall(s) was found, the seeds were displaced toward the edges of the fruit, and these seeds were about 75% smaller and much softer than the seeds of healthy fruits. In addition, the mesocarp/endocarp was much harder, whereas in uninfected fruits this tissue becomes fleshy.
Phenology of the plant and the galls
Trees at Zurquí had fruits during the entire year, but there were more trees with fruits between May and August, and a reduction between December and March (c 2 = 21.69, 11 df, p <0.05). Flower production varied during the year (c 2 = 51.62, 11 df, p <0.001), with trees producing more flowers from April to August. In Escazú most trees produced fruits just once during the year, and only three individuals had fruits most of the year. The number of trees with flowers (c 2 = 151.51, 14 df, p <0.001) and fruits (c 2 = 72.58, 14 df, p <0.001) varied between months of the year. There were more trees with flowers from February to March, while in September there were none; fruiting began in April and May and finished in December. The number of trees with and without galls varied throughout the year (Kolmogorv–Smirnov: Dmax = 0.54, p <0.05). At both sites there were more trees with galled fruits during the rainy season (August to November) than during the dry season (t = − 3,22, 22 df, p = 0.004) ( Figure 2 View Figure 2 ). Trees at Zurquí had galls throughout the year, although the number of galls diminished between December and April.
At Zurquí there was temporal variation in the number of larvae and adults (Kolmogorov–Smirnov: Dmax = 0.70, p <0.025), but when pupae were included in the analysis there was no variation (Kolmogorov–Smirnov: Dmax = 0.60, p> 0.100). Larvae occurred twice during the year, from January to April and from July to August. Young larvae were more abundant in January and July, middle-aged larvae in February and August, and mature larvae in March and August. Most pupae were found from March to April, and in August. There were more adults in April (n = 203), and from September to November (n = 467). There was also temporal variation in the number of larvae, pupae and adults at Escazú (Kolmogorov–Smirnov: Dmax = 0.54, p <0.05). At this site larvae were most abundant from August to November and all larval stages had a peak of abundance in October. Although the sample of pupae was very small, they showed the same temporal pattern in abundance. Adults were most abundant from August to October.
Effects of galls on the host plant
A total of 7791 fruits were examined, 4421 from Zurquí and 3370 from Escazú. Of the total number of fruits examined from Escazú only 15% had galls, whereas at Zurquí 58% had galls (c 2 = 1448.21, 1 df, p <0.001). At the latter site a greater number of trees with fruits had galls (t = 5.31, n = 20, p <0.001), but the fruits at Escazú had more galls per fruit (2.85 ± 1.94) than those at Zurquí (2.41 ± 1.65) (t = 5.74, 3079 df, p <0.001).
At both sites fruits in development stages 1 and 2 had more galls: 98% of the total fruits examined with galls at Zurquí and 97% at Escazú; only 1.75% of the galled fruits at Zurquí and 2.5% at Escazú were in stagese 3 and 4 (Zurquí c 2 = 112.31, 3 df, p <0.001; Escazú c 2 = 194.31, 3 df, p <0.001). Fruits in stages 1 and 2 in general (galled and ungalled) were more abundant than those in stages 3 and 4 (Zurquí c 2 = 3745.10, 3 df, p <0.001; Escazú c 2 = 2854.94, 3 df, p <0.001) ( Figure 3 View Figure 3 ). The average number of galls per fruit was greatest in stage 2, and these were also the fruits that had the greatest maximum number of galls per fruit (14 galls) (c 2 = 931.11, 12 df, p <0.001). Maximum numbers of galls per fruit were seven galls for stage 1, eight galls for stage 3 and six galls for stage 4.
The nested analysis of variance showed that the number of galls per fruit affected the volume of fruits in all developmental stages (stage 1 F = 174.81, 1 df, p <0.001, stage 2 F = 434.53, 1 df, p <0.001, stages 3 + 4 F = 14.40, 1 df, p = 0.001). The volume of the fruits was significantly greater when they had galls than when they lacked galls (stage 1 t = − 17.79, 2438 df, p <0.001, stage 2 t = − 5.01, 4648 df, p <0.001, stage 3 t = − 5.13, 257 df, p <0.001), except in stage 4 fruits (t = 1.44, 440 df, p = 0.15) ( Figure 4 View Figure 4 ). The volume of stage 1 galled fruits did not vary between the two sites (F = 0.86, 1 df, p = 0.355), but did vary among the individuals within each site (F = 6.68, 36 df, p <0.001). The volume of the stage 1 galled fruits varied from 0.14 to 1.37 cm. The volume of galled fruits in stages 2, 3 and 4 varied within individual trees at each site (stage 2 F = 13.12, 49 df, p <0.001, stages 3 and 4 F = 4.02, 17 df, p <0.001). Stage 2 fruits varied from 0.29 to 6.09 cm, and those in stages 3 + 4 varied from 0.44 to 2.39 cm. Average fruit volume was greater at Escazú (stage 2 1.21 ± 0.55, stage 3 1.61 ± 0.42, stage 4 1.27 ± 0.2) than at Zurquí (stage 2 0.89 ± 0.31, stage 3 1.02 ± 0.38, stage 4 1.03 ± 0.36) (stage 2 F = 11.91, 1 df, p <0.001, stages 3 + 4 F = 6.32, 1 df, p = 0.016).
Ungalled stage 1 fruits at Escazú had a greater average volume than did those at Zurquí (site F = 11.66, 1 df, p = 0.01). There was little variation among trees in the two populations (trees: F = 2.4, 51 df, p <0.001) [Zurquí coefficient of variation (CV) = 0.27, Escazú CV = 0.22]. Ungalled fruits in stages 2, 3 and 4 showed differences in volume among trees at each site (stage 2 F = 2.29, 48 df, p <0.001, stages 3 + 4 F = 2.66, 43 df, p <0.001), but no differences were found between sites (stage 2 F = 1.14, 1 df, p = 0.286, stages 3 + 4 F = 1.42, 1 df, p = 0.283).
The average number of seeds per fruit diminishes significantly when fruits (of all stages) have galls (t = − 7.73, 476 df, p <0.001) ( Figure 5 View Figure 5 ). The average number of seeds in galled fruits in stages 1 and 2 was similar between the two sites (stage 1 F = 2.51, 1 df, p = 0.117, stage 2 F = 4.04, 1 df, p = 0.05), but not among trees (stage1 F = 2.96, 15 df p <0.001, stage 2 F = 4.24, 34 df, p <0.001). There was greater variation in seed number in stage 1 fruits within trees at Zurquí (Zurquí: CV = 0.58, Escazú: CV = 0.34). For stages 3 and 4, the sample size was insufficient to examine differences between sites and individuals.
No known copyright restrictions apply. See Agosti, D., Egloff, W., 2009. Taxonomic information exchange and copyright: the Plazi approach. BMC Research Notes 2009, 2:53 for further explanation.