We investigated female black flies on or near (within 3.2 km of the property boundary) Necedah National Wildlife Refuge (NNWR), Juneau County, WI, U.S.A. (Latitude: 44.149; Longitude: –90.183). We surveyed larval and pupal black flies on the NNWR and for 10 km beyond its boundary, a distance likely to include typical dispersal distances of the females of most species (Adler et al. 2004). The study area is characterized by flat topography. NNWR has more than 200 km of drainage ditches. Two rivers, the Yellow River on the east and Beaver Creek on the west, flow north to south along the NNWR boundary, and the Lemonweir River lies south of the NNWR (Figure 1). The entire study area, including the streams and rivers, has a substrate of coarse sand. Surveys of adult black flies were conducted in shallow meadows (water depth about 60–75 cm) used by nesting whooping cranes. The meadows were dominated by sedges (Carex spp.) and bulrushes (Scirpus spp.), with patches of willow (Salix spp.) in drier areas, and were located largely in open prairie areas maintained with prescribed burning and mowing since the 1940s. The only exceptions were sedge meadows in wooded areas with black oak (Quercus velutina), jack pine (Pinus banksiana), and aspen (Populus grandidentata and P. tremuloides).
Figure 1. Location of whooping crane (Grus americana) nests, artificial nests, and carbon-dioxide (CDC) traps in and around Necedah National Wildlife Refuge, Juneau County, WI, U.S.A.
Download figure to PowerPoint
Experimental procedure and data analysis
Approval to work in and around active whooping crane nests and to work with eggs of red-crowned cranes (Grus japanensis) and Siberian cranes (Grus leucogeranus) and body parts of whooping cranes was covered under an Endangered Species Recovery Permit (#TE048806-1). The research also was covered by a Migratory Bird Permit (#MB09144A-1).
To determine the local species pool of black flies, we surveyed all sites accessible by vehicle, both on the NNWR (representing virtually all flowing water on the Refuge) and off the Refuge (representing all flows 2 m or more in width) for 20–40 min each. In total, we sampled 57 sites on the Refuge and 20 off the Refuge at various dates from 30 March to 6 May in 2009–2011. Larvae and pupae were collected with forceps from all available substrates (e.g., trailing vegetation, leaf packs) at each site and fixed in three changes of 1:3 acetic ethanol. Identifications were made morphologically and cytogenetically, following procedures and identification keys of Adler et al. (2004).
Host-seeking black flies were collected in 2009 and 2011 with seven Centers for Disease Control (CDC) traps (CDC MiniLight Trap, Bioquip Products, Rancho Dominguez, CA, U.S.A.) established randomly in suitable whooping crane nesting habitats (Figure 1). Suitable sites were based on nest locations on NNWR from 2005 to 2008. We used ArcMap 9.3 (ESRI Inc., Redlands, CA, U.S.A.) to model suitable whooping crane nesting habitats and to select trap locations randomly. Trap locations were the same in 2009 and 2011. The distance from the traps to the nearest whooping crane nest was 1,358.7±215.8 m and 756±139.3 m in 2009 and 2011, respectively.
Traps were operated daily with the light deactivated and individually baited with about 1.6 kg of dry ice each, from 14 April to 6 June 2009 and from 11 April to 17 June 2011. The only exceptions in 2009 were days when wind or rain would interfere with trapping operations. In 2011, the traps were deployed one or two days per week, weather permitting. Thus, traps were deployed on 53 days in 2009 and on 17 days in 2011. We had a total of 343 trap nights in 2009, with 36 trap malfunctions, and 117 trap nights in 2011, with only two trap malfunctions. In both years, trapping began before the emergence of black flies. Because trap height can influence collecting efficacy (Russell and Hunter 2005, Swanson et al. 2012), we placed the traps about 150 cm above the water, the approximate height of a whooping crane's head while on a nest.
We counted female black flies at whooping crane nests by combining the results of three methods: 1) glueboards, 2) high-resolution images, and 3) specimens drowned in the contents of damaged or hatched eggs. These data were collected once, upon completion of a whooping crane nesting effort (i.e., a nest desertion or a successful hatching). We deployed 40-mm × 65-mm glueboards (Professional Pest Control, Columbus, GA) on top of the heads of sandhill crane decoys (Model Q1600, Carry-Lite Decoys, Ft. Smith, AR) painted to resemble whooping cranes. Glueboards were exposed for 5 min, following the procedure of Weinandt4. To avoid collecting black flies attracted to humans, we remained at least 25 m from the decoy during the 5-min exposure. After the glueboard sampling, we removed the decoy and obtained an image of the nest from 2 m, with a superfine-resolution (2816 × 2112 pixels) camera. Finally, we collected any damaged or hatched eggs from the nest. We summed results of all three counts by species to give a count per nest. For black flies counted on high-resolution images, we applied the species ratio from the glueboard and egg samples. When we were able to collect a glueboard sample but not a high-resolution image or vice versa, we limited black fly results to presence/absence information.
Black flies from carbon-dioxide traps were euthanized by freezing for 2 h. Those on glueboards were removed with xylene. Individuals from all samples were fixed in 95% ethanol and all specimens were counted. All glueboard and egg samples were identified to species, as were all carbon-dioxide trap samples with fewer than 120 individuals. For samples from carbon-dioxide traps with more than 120 black flies, we counted the entire sample and randomly selected 120–200 specimens for identification using a petri dish on a grid. Specimens were boiled in 10% potassium hydroxide (KOH) to remove soft tissues, washed twice in distilled water, and identified by microscopic examination of genitalic features (Adler et al. 2004). The same method was used for specimens from glueboards, with the addition of 10% mineral spirits after boiling in KOH. Representative specimens of all species were deposited in the Clemson University Arthropod Collection, Clemson, SC.
To evaluate the performance of carbon-dioxide traps for describing variability in black fly populations at whooping crane nests, we used the following comparisons: 1) black flies at a nest vs the mean number captured in all seven carbon-dioxide traps and the number captured in the carbon-dioxide trap nearest the nest, 2) positive and negative indices (presence/absence) for black flies at a nest vs seven carbon-dioxide traps and the nearest carbon-dioxide trap, and 3) correlations between the number of black flies (when present) at nests and the mean number from all seven carbon-dioxide traps and from the nearest carbon-dioxide trap. We limited these comparisons to occasions when we had results from carbon-dioxide traps on the same, previous, or following day of a nest visit. All data were log (n+1) transformed before analyses with paired t-tests (Muturi et al. 2007, Smallegange et al. 2010, Chen et al. 2011) and Spearman's rank correlation (Comtois and Berteaux 2005, Bisevac et al. 2009, Swengel and Swengel 2011).
We constructed nine artificial nests in 2010 to evaluate their performance in attracting black flies, compared with real whooping crane nests (Figure 1). To mimic whooping crane nests, we piled approximately 45 cm of bulrush, the primary material used by whooping cranes (Allen 1952), in a woven fashion to cover a 76 × 76-cm floating piece of oriented strand board secured with wire to two submerged stakes. In 2009, we found that black flies were attracted to whooping crane eggs; we, therefore, baited each artificial nest with one randomly assigned red-crowned crane or Siberian crane egg. The eggs were not incubated by adult cranes and were chilled before deployment. To protect the eggs from scavengers, we constructed exclosures (30 cm long × 15 cm wide × 20 cm high) consisting of a wooden frame (2.5 × 2.5 cm) covered with chicken wire. We obtained a high-resolution image of each baited nest and collected images for three subsequent days, from 14 to 17 April. Nest-visiting species of black flies began emerging from their breeding habitat on 9 April 2010. By 14 April, we had visited several nests and confirmed the presence of black flies. The performance of artificial nests was evaluated by comparing fly counts from high-resolution images with those from real nests.
Birds use uropygial gland secretions to keep their feathers clean and waterproof. They transfer the secretions by rubbing their head against the uropygial gland at the base of the tail and then rubbing the saturated head feathers against feathers on the rest of the body. Because uropygial gland secretions attract some species of black flies (Fallis and Smith 1964, Bennett et al. 1972), we measured their efficacy by comparing glueboard collections from unbaited decoys with those from decoys baited with real whooping crane wings. Decoys were deployed randomly on nests. The wings were stored in a freezer when not in use and transported to the nests in garbage bags on ice in a cooler. Fly counts from baited and unbaited decoys were compared using a paired t-test on logarithmically (n+1) transformed data (Muturi et al. 2007, Smallegange et al. 2010, Chen et al. 2011). All statistical analyses for all experiments were done with SAS 9.1 (SAS Institute Inc., Cary, NC, U.S.A.).