We found a diverse sawfly fauna of 176 species in 66 genera in 10 families in GWMP which consume at least 57 genera of angiosperms, ferns, gymnosperms, and horsetails (Table 1). These sawflies include a rarely-collected species (Kerita fidala Ross) recorded as a leafminer of Mertensia virginica (L.) Pers. ex Link (Boraginaceae) and an unidentified Caliroa sp. which consumes Staphylea trifolia L (Staphyleaceae) and is still known only from larvae. Both sawfly species were collected in the floodplain near Turkey Run. The record of K. fidala is the first for Virginia (Smith 2009). Nine of the collected species are alien in North America. Ninety-five percent of the reported plant host genera of the sawfly species that we caught occurred in all three of the parks.

We found 69 species in DMWP, 134 in GFP, and 115 in TRP. Twelve species were unique to DMWP, 43 to GFP, and 22 to TRP. Forty species were found in all three parks. Coefficients of Community (CCs) for GWMP ranged from 0.49 through 0.68 (Table 2). Dyke Marsh Wildlife Preserve was more similar to TRP than GFP, and GFP and TRP were more similar to one another than to DMWP.

As a group, GWMP sawflies flew from early March through mid-November. With regard to species richness and month, more species occurred in April through June than in other months, with species numbers peaking in May in each of the three parks and for all parks combined (Fig. 2). In DMWP, the number of individuals peaked in early June 1998 and early May 1999.

Figure 2. 

Number of sawfly species versus month in the GWMP, Virginia. Diamonds represent sawfly numbers in DMWP; squares, GFP; triangles, TRP; and circles, all three parks combined.

Of the 176 species collected, 138 have a single emergence period in the spring, March through June, indicting a single generation a year. Another 12 species apparently with only a single generation fly only from mid-summer through October, with only 3, both species of Metallus and Tremex, occurring as late as September and October. The remaining 26 species occur through much of March or April through September and are probably multivoltine. Species of two genera, Allantus and Taxonus are apparently bivoltine, with a large emergence in the spring and a small emergence in late summer or early autumn.

Of the 695 sawfly specimens trapped in DMWP 73% were from 1998, and 27% were from 1999. In both years combined, 42% were in the ecotone; 39%, forest; and 19%, marsh (Fig. 3). A plot of frequency of specimens versus species (Fig. 4) indicates that only a few species were present in large numbers; 71% of the 69 species had 9 or fewer individuals in our sample.

Figure 3. 

Sawfly abundances (mean per day per sampling period) varied over their flight periods in 1998 and 1999 in Dyke Marsh Wildlife Preserve, Virginia. Diamonds represent sawfly numbers in the ecotone; squares, forest; triangles, marsh; and circles, all three habitats combined. In 1998, sampling interval A1 was from 12–19 April; A2, 19–28 April; M1, 28 April – 10 May; M2, 10–17 May; M3, 17–28 May; J1, 28 May – 6 June; J1, 6–14 June; J3, 14–24 June; J1, 24 June – 7 July; J2, 7–19 July; 1J3, 19–30 July; A1, 30 July – 9 August; A2, 9–12 August; A3, 12–28 August; S1, 28 August – 11 September; S2, 11–26 September; O1, and 26 September – 11 October. No sawflies were in the samples from 11–26 October and 26 October – 6 November; 6–22 November, 22 November – 6 December, and 6–20 December, which are not shown on this graph. In 1999, sampling interval F1 was from 24 January – 15 February; M1; 15 February – 6 March; M2, 6–20 March; A1, 20 March – 3 April; A2, 3–11 April; A3, 11–25 April; M1, 25 April – 8 May; M2, 8–23 May; J1, 23 May – 6 June; J2, 6–20 June; J1, 20 June – 2 July; J2, 2–18 July; J3, 18–23 July; A1, 23 July – 8 August; A2, 8–15 August; A3, 15–29 August; S1, 29 August – 12 September; S2, 12–26 September; O1, 26 September – 11 October; O2, 11–24 October; 16, N1, 24 October – 8 November; and N2, 8–21 November. There were no sawflies in samples J1, 20 December – 3 January; J2; 3–24 January; D1, 21 November – 5 December; and D2, 5–19 December which are not shown on this graph.

Figure 4. 

Number of species versus number of individuals captured in the DMWP sample. In addition, one species had 58, one had 84, and one had 251 individuals which are not shown on this graph to conserve space.

We found 176 sawfly species in the GWMP; however, this biotically-rich location might harbor over 200 species. A similar study of species abundance and diversity in two oak-pine forests in Virginia and West Virginia using Malaise traps estimated that 81% of the actual species present were captured over a five-year period (Braud et al. 2003, Strazanac et al. 2003). We expected to find more diprionid species and species of conifer-feeding siricids. We found three Acordulecera species, but more are expected once taxonomic problems are resolved. There are several reasons why we did not find more species in our survey. Malaise traps are biased toward strong flying species such as Taxonus and Macrophya; species that are weak flyers or stay close to their host plant are not often collected in traps. Some mid-Atlantic species are rare or not expected in GWPM. Immature stages of many species are very difficult to find in nature. Adults of some species tend to be in places such as near their hosts or in treetops where we did not site our traps. More intense sampling in already-sampled habitats and sampling of additional unsampled GWMP habitats could obtain more species. Unsampled GWMP sites include areas in Montgomery Co., Maryland, and the District of Columbia, as well as sites with Pinus and Salix, the floodplain forest of Theodore Roosevelt Island, the Clara Barton Parkway in Maryland, and the George Washington Memorial Parkway in Virginia, where these roads border and run through the GWMP.

The GWMP sawfly fauna is 50% of the 351 sawfly species recorded for the entire state of Virginia (Smith 2006, 2013). In comparison, Smith and Barrows (1987) found 85 species in samples from less natural Maryland habitats using two sets of Malaise traps 0.8 and 1.5 km from TRP. Seventy percent of these species are also in our samples from the three GWMP parks. There are 91 recorded species for Plummers Island, Maryland, located across the Potomac River from TRP (Smith 2008). In another Virginia Coastal Plain site, D. R. Smith (unpublished) found 175 species in Essex Co. based on 10 years of sampling. In other Virginia Piedmont sites, he found about 200 species in the University of Virginia Blandy Experimental Farm and State Arboretum of Virginia, Clarke Co., based on 6 years of sampling; 175 species in Bull Run Mountains Conservancy, Prince William Co., in 3 years; and 125 species in a his suburban yard in Fairfax Co. in 33 years.

Flight times and numbers of adult sawflies can vary from year to year as occurred in the DMWP samples. Factors including distribution of host plants, adult foods, and mates; drying winds; natural enemies; soil moisture and temperature; and weather affect adult abundances, distributions, and flight periods (Wallace and McNeal 1966), subjects not yet studied in DMWP.

The causes of the sawfly abundance and distribution in DMWP are not yet studied. Our plot of frequency of specimens versus species (Fig. 4) shows that the majority of species were not common in our samples. Smith and Barrows (1987) found a similar relationship in their sample from a less natural area and a yard in the Washington, D.C., area. Low numbers of specimens of many species may have occurred in our sample because of factors such as a species’ being rare, rare individual's of more common species straying into our study site from elsewhere, and a species’ flight habit precluding our trapping the species. In DMWP, adult sawflies were more common in 1998 than in 1999 based on trap samples, as also occurred in other DMWP insects such as fireflies (Barrows et al. 2008) and mecopterans (Barrows and Flint 2009). In contrast, DMWP rhopalosomatids (Barrows 2013) and sialids (Barrows et al. 2005) were more frequently captured in 1999 than in 1998.

In summary, we found a rich fauna of 176 sawfly species which feed on many plant genera in the GWMP, and as a group, the sawflies flew from early March through November. Sawfly species in DMWP were most common in its ecotone, followed by its forest and marsh, and these species greatly varied in abundance. In this time of worrisome, rapid global change, threats to the GWMP include air, soil, and water pollution; many alien, invasive species; flooding; and erosion (Litwin et al. 2013, pers. obs.). In fact, DMWP is losing about 0.6–0.8 hectare per year due to erosion. The National Park Service may restore the marsh to some extent in this decade. Increased marshland from restoration could change the mean population sizes of some the DMWP sawfly species. Since Symphyta is a species-rich, GWMP taxon, it is an appropriate one for monitoring GWMP’s health in forthcoming years. This large, wasp suborder is an understudied animal taxon, and myriad aspects of the biologies of GWMP and other species are ripe for investigation.