Stenodiplosis spartinae Gagné, 2015

López, J. Manuel Perilla, Johnson, Paul J., Gagné, Raymond J. & Boe, Arvid, 2015, A new species of Stenodiplosis (Diptera: Cecidomyiidae) on Spartina grasses (Poaceae) with notes on its biology and its parasitoid Tetrastichus bromi (Hymenoptera: Eulophidae), Zootaxa 4057 (1), pp. 115-124 : 117-122

Stenodiplosis spartinae Gagné , new species

( Figs. 1–8, 10–11 , 14–15 , 18 )

Description. Adult. Head: Eyes connate, 3–4 facets long at vertex, facets circular, slightly farther apart laterodorsally than elsewhere. Frons with 9–20 setae. Mouthparts ( Figs. 2–3 ): labella spheroid, pointed apically, setulose, each with 3–5 lateral setae; palpus 1- or 2-segmented; when 2-segmented, second palpal segment less than half length of first. Male antennal flagellomeres ( Fig. 4 ) with short internode and neck. Female flagellomeres cylindrical, without necks.

Thorax: Wing: hyaline; length in male, 1.9–2.4 (n=10), in female, 2.2–2.6 (n=10); R5 with slight curve, joining C near wing apex; CuA forked. Acropods ( Fig. 1 ): about half as long as fifth tarsomeres; claws untoothed, curved beyond midlength; empodia slightly longer than claws; pulvilli ca. 1/3 length of empodia.

Male abdomen: Tergites rectangular, all lacking scales, first through fifth with posterior setae in a mostly single, medially discontinuous row, sixth and seventh with mostly double row of posterior setae nearly or wholly continuous medially, eighth pigmented only on anterior half, without vestiture except for anterior pair of trichoid sensilla. Sternites rectangular with single to double row of posterior setae, scattered setae and setiform scales near midlength, and with anterior pair of trichoid sensilla present only on eighth sternite. Terminalia ( Figs. 5–8 ): cercus ellipsoid, with a few setae along distal edge; hypoproct deeply notched apically, the resulting cylindrical lobes broadly rounded apically, each with a long apical seta and several shorter ventral setae; gonocoxite bulbous; gonostylus broadest near midlength, completely setulose, with several short, scattered setae; apodeme broad, anterior margin deeply concave; aedeagus tapered, about as long as hypoproct, with pair of sensilla on each side. Female abdomen ( Figs. 10–11 ): Tergites first through seventh as in male, eighth tergite square, about half width of seventh, with single row of short posterior setae and pair of anterior trichoid sensilla. Sternites as for male except eighth sternite evanescent and marked only by presence of anterior pair of trichoid sensilla. Cerci elongate, gradually tapered, asetulose; each with 2–3 comparatively long setae near base, a pair of stiff, thick setae subapically ( Fig. 11 , arrow), and several smaller setae along length and at tip.

Type Specimens. HOLOTYPE, male; labelled from Spartina pectinata, Aurora Research Farm, 2.7 km NNE of Aurora, el. 505 m, 44.306N, 96.671W, Brookings Co. SD, coll. VII-15-2013, em. VII-18/20 -2013, P.J. Johnson, (deposited in USNM).

PARATYPES (all from Spartina pectinata ): 2 males, 5 females, same data as holotype (1♂ 2 ♀ SDSU, 1♂ 3 ♀ USNM); 3 males, 4 females, Felt Family Farm, 6.3 km N of Brookings, 44.369N, 96.795W, Brookings Co., SD, VIII-16-2012, P.J. Johnson (1♂ SDSU, 2♂ 4 ♀ USNM); 6 females, 12 Km NNE of Coleman, 44.087N, 96.799W, el. 511 m, Moody Co., SD, coll. VII-23-2014, em. VIII-1 /6-2014, J.M. Perilla.

Additional specimens. 3 males, 3 females, from inflorescences of Spartina alterniflora, Ocracoke I. , 35.1128°N, 75.9758°W, Hyde Co., NC, coll. IX-14-1968, M.T. Mooring; and 2 males, 2 females, from inflorescences of Spartina alterniflora, Pine Knoll Shores , 34.6958°N, 76.8178°W, Carteret Co., NC, N.H. Newton. All deposited in the USNM.

Etymology. The epithet “ spartinae ” is derived from Spartina , the genus of the host for this new species of gall midge.

Distribution. This species is known from eastern South Dakota and coastal North Carolina.

Remarks. Of the seven species of Stenodiplosis now known from North America, four have palpi with four segments, and three, including the new species, have palpi with one or two segments. The three species with reduced palpal segments, S. spartinae , S. albescentis and S. phragmicola , are all native Nearctic species. Stenodiplosis spartinae shows variation in having either one ( Fig. 3 ) or two segments ( Fig. 2 ), unlike S. albescentis and S. phragmicola , which have palpi with two equally long segments. This may occur even on the same specimen ( Fig. 2 ) and, when two-segmented, the second palpal segment is never more than half the length of the first segment. Stenodiplosis spartinae further differs from S. phragmicola in the ventrally setose and more cylindrical lobes of the male hypoproct ( Fig. 7 ) and in the less strongly tapered female cerci, with the subapical pair of setae longer than the width of both cerci at the point where the cerci are inserted ( Fig. 11 ). In S. phragmicola the lobes of the male hypoproct ( Fig. 9 ) are flatter and more triangular and have no ventral setae, and the female cerci are more strongly tapered and their subapical pair of setae are shorter than the width of both cerci ( Fig. 12 ). Stenodiplosis albescentis differs from S. spartinae in the length of the acropods that are as long, rather than half as long, as the fifth tarsal segment, and in the female cerci that have longer and pointed subapical setae ( Fig. 13 ).

Biology. Stenodiplosis spartinae is associated with the inflorescence of its host Spartina pectinata . This is the first identified gall midge reported from S. pectinata and from an inflorescence of a Spartina species, although the “dipteran larval seed parasites” of Bertness et al. (1987) on S. alterniflora and S. patens were most likely gall midges. A reproductive culm, or tiller, of S. pectinata may exceed 3 m in length under optimum growing conditions, and the inflorescence normally projects well above the canopy height of the leaves. An individual spike of the inflorescence has from 11–25 spikelet pairs arranged in two appressed rows along a rachis, with upwards of 30 spikes per panicle. Spikelets of S. pectinata are deciduous and fall to the ground in late autumn through winter, where they may be consumed by various invertebrates and small vertebrates. Many of the spikelets remain mostly intact and surviving caryopses may germinate the following spring, or in subsequent years.

Stenodiplosis spartinae adults ( Fig. 14 ) were found active from mid-July through August, when prairie cordgrass blooms regionally. The flies appeared and began ovipositing as the inflorescence emerged from the sheath of the flag leaf, with peak oviposition at about 75% extrusion of the inflorescence. Within a few days of first observed oviposition, first instars were found feeding on pre-anthesis unfertilized ovules. During larval feeding no external change in color or shape of spikelets was observed. Oviposition continued post-anthesis, but progressively diminished and was eventually restricted to only late developing inflorescences.

Adults reared in laboratory conditions emerged from 31 July to 0 1 September in 2014, with 15 males and 11 females (sex ratio 0.57:0.43) emerging from 849 spikelets. However, infestation rates could not be accurately established in the field, due to the complementary spikelet and floral structure damage inflicted by Aethes spartinana (Barnes and McDunnough) ( Lepidoptera : Tortricidae ), Melanoplus bivittatus (Say) and M. femurrubrum (De Geer) ( Orthoptera : Acrididae ), and Diabrotica barberi Smith and Lawrence ( Coleoptera : Chrysomelidae ). Each of these species prevents pollination of florets through their feeding damage.

Adult gall midge activity was observed in the field from approximately 09:30–20:00 hrs from mid-July through August. Beginning on 16 July 2014, individual females were observed to oviposit repeatedly in florets before, during, and shortly after initiation of anthesis. Female activity occurred from mid-morning to evening, with the greatest activity occurring on overcast days from 16:00–18:00 hrs, when the wind was <14 km /h. If a gravid female landed on the basal areas of a spike, she would oviposit in several florets while moving apicad. Oviposition was observed near the apex of the floret where the pliable ovipositor was inserted under the apical edges of either the palea or lemma. A female that alights at or near the apex of an upright spike walks circularly near the apex while ovipositing in several of the terminal spikelets.

By the first week of August, some orange-colored third instars were found leaving the spikelets, moving down the rachis, and presumably dropping into the duff and soil at the base of the host plant. These larvae may be endoparasitized or may be entering the soil before a winter diapause. Third instars were not found on plants by the first week of September and we were unable to verify that larvae remained within the deciduous spikelets in the field during the winter.

The initial emergence of adults in mid-July was evidently from overwintering larvae. Field observations of activity patterns and the detection of two midsummer emergence peaks, followed by similar activity peaks by parasitoids, suggests that there may be development periods of two to three weeks each during July and August, allowing at least two, or possibly three, generations per year. It is likely that there is some degree of overlap of these generations.

The larva of S. spartinae ( Fig. 15 ) feeds on the ovule and developing caryopsis during its early clear fluid, milky and soft dough developmental stages, when the pericarp is soft and before the nutrient fluids solidify into endosperm. Larval feeding kills an ovule, and causes collapse, shrinking, and death of the developing caryopsis ( Fig. 19 , right). This feeding is similar to damage described by Barnes (1956) for S. sorghicola on sorghum. In Spartina pectinata the lengths of infested caryopses were approximately 12–25% of undamaged, fully developed caryopses. During larval feeding we observed that the head of the larva is pressed against the ovule or soft seed coat, and the larva either consumed the ovule or embryo, or entered and completed its development within the immature caryopsis, the larva of S. spartinae lacks oral structures for rasping or boring through tissues. A similar observation on feeding was reported by Agafanova (1962) for S. bromicola on Bromus inermis .

In dissected florets that were not pollinated, the ovule remained small and pallid, and was occupied by visibly weak second instars that appeared unable to complete their development. Apparently, active metabolic activity of the plant embryo is necessary for complete larval development. Third instars were found only in florets with fully consumed ovules, or those that were pollinated, and larvae were in the developing caryopses. Peak larval feeding was observed from late July to mid-August.

Prasifka et al. (2012) found that the caterpillars of Aethes spartinana were “seed predators” and concentrated on the central spikelets of Spartina pectinata , with a 38% loss of seed production due to this insect. Boe et al. (2013) did not specifically mention a predator, but considered insect predation a significant factor in low seed set in native stands of Spartina pectinata . We found no evidence of co-occupation of a spikelet by the Stenodiplosis larva and Aethes spartiniana caterpillar. This suggests that the S. spartinae larva feeding on ovules and early development caryopses, versus later stage development caryopses for A. spartinana caterpillars, segregate these species phenologically, thereby avoiding direct competition. Ovule and early caryopsis feeding by S. spartinae suggests that this gall midge may contribute significantly to natural and commercial seed loss in Spartina pectinata .

During activity periods of S. spartinae a parasitoid determined by Z. Yefremova, Tel Aviv University (in litt.) as the Palaearctic Tetrastichus bromi Kostyukov ( Fig. 16 ) was found commonly on the spikes of each inflorescence and was reared from the gall midge. A transfer of this species to Aprostocetus Westwood is pending (Z. Yefremova in litt.). This wasp was active on Spartina pectinata with S. spartinae from the time of emergence of the inflorescence in July until the first week of September, with up to eight adult females often present simultaneously on early spikes when S. spartinae was ovipositing. It was common to find both the wasp and the gall midge simultaneously on a given spike. The wasp actively searched the spike and individual spikelets, frequently stopping to oviposit into the base of a floret. The T. bromi larva ( Fig. 17 ) is ectoparasitic ( Fig. 18 ) suggesting that oviposition is external to the gall midge larva.

The association of Tetrastichus bromi with this gall midge is notable because of the wasp’s host association with other grasses and gall midge species, and its introduction from European Russia. This wasp was formerly reported as Aprostocetus nebraskensis (Girault) [in Tetrastichus Haliday ( Burks 1979, Noyes 2014)] and under the older binomial is transcontinental and was reported from New Brunswick to British Columbia in Canada, and from Maine and South Carolina to New Mexico and Oregon in the United States ( Burks 1943, 1979; LaSalle 1993). Aprostocetus “ nebraskensis ” was historically placed in Tetrastichus , often without determination to species. In South Dakota, Boe and McDaniel (1990) confirmed Aprostocetus nebraskensis as a parasitoid of S. wattsi on Andropogon gerardii . We also observed this wasp, sometimes abundantly, on Alopecurus pratensis , Andropogon gerardii , Bromus inermis , Schizachyrium scoparium , and Sorghastrum nutans . Parasitoids reared from S. spartinae as well as S. bromicola , S. geniculati , and S. wattsi were determined as a combination of T. bromi , A. nebraskensis , and three additional unnamed and apparently undescribed species of Aprostocetus . These form what appears to be a guild of parasitoids specialized on a set of native and introduced grass-associated gall midges. Multiple species of Aprostocetus are also known to parasitize S. sorghicola (e.g., Baxendale et al. 1983, Kausalya et al. 1997, Nwanze et al. 1998).

A “ Tetrastichus ” sp. parasitizing S. bromicola on cultivated selections of Bromus inermis in Nebraska, Oklahoma, and Canada, was reported by Ahring (1965), Neiman and Manglitz (1972), Carter, et al. (1988), Vogel and Manglitz (1989), and Otfinowski et al. (2007). Elsewhere this wasp may not be restricted to grass-feeding gall midges ( Guppy 1961, Moser 1965), if these identifications were correct. Neiman and Manglitz (1972) commented that it “is probable” that the Tetrastichus sp. parasitizing S. bromicola in Nebraska is the same as one reported by Agafonova (1962) from Kursk, Russia, and probably described as T. bromi Kostyukov (1978) . The question then arises as to the source and distribution of T. bromi in North America.

Bromus inermis was first documented as intentionally introduced to North America and into California around 1880 when seeds from Hungary were sold in 4 oz packets (Dwinelle 1884). Lyon (1899) reported that this grass was first cultivated at the Nebraska Experiment Station in 1897. In 1898 N.E. Hansen, working for the U.S. Department of Agriculture Office of Foreign Seed and Plant Introduction, was authorized to purchase and import 12 tons of B. inermis seed from the Penza Province of Russia to the USA, with the shipment arriving in February 1899 ( Newell and Keim 1943, Newell 1973, Dunn 1985, Otfinowski et al. 2007). Of this amount over 375 pounds of seed was sent to the Nebraska experiment station for testing and redistribution, with the remainder sent to at least 41 other states ( Dunn 1985). The original description of Aprostocetus nebraskensis (Girault) was published during the subsequent active period of Bromus inermis varietal selection and testing in and around Lincoln, Nebraska, from which its association as a common gall midge parasitoid began. Stenodiplosis bromicola diapauses within bromegrass spikelets and may remain in diapause for two years ( Agafonova 1962, Neiman and Manglitz 1972), supporting the suggestion that this gall midge may have been introduced in S. bromicola infested seed. This gall midge was first reported in the United States from Lancaster Co., Nebraska ( Gagné 1966), which includes the Lincoln area. This introduction of S. bromicola and T. bromi parallels the probable mode of introduction of seed of Russian genotypes of Euphorbia esula (leafy spurge) with the Hansen bromegrass seed purchase ( Dunn 1985).

We found “ Aprostocetus nebraskensis ”, now determined as Tetrastichus bromi , on or reared from Alopecurus pratensis and Bromus inermis , both Eurasian cool-season grasses. The true A. nebraskensis seems restricted to S. wattsi on the native warm-season grasses Andropogon gerardii , Schizachyrium scoparium , and Sorghastrum nutans . The host gall midges, S. bromicola and S. geniculati in the introduced grasses are also Eurasian species. It appears that T. bromi is an oligophagous Eurasian introduction specialized on Stenodiplosis larvae that then transferred to North American native Stenodiplosis spp., including S. spartinae . The community dynamics impact of this wasp host shifting onto the native Stenodiplosis parasitoid guild remains to be explored.

Figures

FIGURES 1 – 13. Stenodiplosis species. Figs. 1 – 8. Stenodiplosis spartinae: 1. Acropod. 2. Palpi and labella (frontal). 3. Onesegmented palpus. 4. Male fourth flagellomere (ventral). 5, Male terminalia (dorsal). 6, Male cerci (dorsal). 7, Aedeagus and hypoproct (ventral). 8, Cerci, hypoproct and aedeagus (lateral). Fig. 9. Stenodiplosis phragmites, male hypoproct (dorsal). Figs. 10 – 11. Stenodiplosis spartinae, 10. female postabdomen, fifth segment to cerci (dorsal). 11. Stenodiplosis spartinae, female cerci (detail, dorsal; arrow indicates subapical pair of setae). Fig. 12. Stenodiplosis phragmites, female cerci (dorsal). Fig. 13. Stenodiplosis albescentis, female cerci (dorsal).

FIGURES 14 – 15. Stenodiplosis spartinae. Fig. 14. Adult Stenodiplosis spartinae on spike of Spartina pectinata; scale bar = 1.0 mm. Fig. 15. Stenodiplosis spartinae third instar (ventral); scale bar = 0.5 mm.

FIGURES 18 – 19. Spartina pectinata floret showing Tetrastichus bromi larva (black arrow) with Stenodiplosis spartinae larva (white arrow) in situ, scale bar = 0.5 mm. Fig. 19. Spartina pectinata caryopsis sheath, fully developed on left; caryopsis sheaths with ovule damaged by Stenodiplosis spartinae on right; scale bar = 1.0 mm.

FIGURES 16 – 17. Adult female Tetrastichus bromi on spikelet of Spartina pectinata; scale bar = 0.5 mm. Fig. 17. Tetrastichus bromi third instar (lateral); scale bar = 0.25 mm.