Mortality of Lepidoptera along Roadways in Central Illinois

We conducted this study to inves tigate the magnitude of roadway mortality of Lepidoptera in central Illinois. To quantify the number and kinds of Lepidoptera killed along roadways, dead adult Lepidoptera were collected, identified, and counted from along 13 roadside transects in the vicinity of ChampaignIU rbana, Illinois , with collections occurring weekly on each transect for six weeks. During the six weeks of this study, 1824 presumably road-killed Lepidoptera were collected. At traffic rates of 1000, 13,500, and 19,700 vehicles per day, more Lepidoptera were collected per 100 m than at other traffic rates. A peak in monarch butterfly mortality may coincide with the timing of their annual migration through the area. Based on these data, the number of Lepidoptera killed along roadways for the entire state of Illinois during one week was estimated at more than 20,000,000 individuals. The number of monarch butterflies killed may have exceeded 500,000 individuals. Our results suggest that increases in traffic rate and speed limit may to a certain extent increase mortality. Additional key words: Butterflies, Danaus plexippus, traffic. Although roadway traffic is known to affect population densities of amphibians (Fahrig et al. 1995), snakes (Bernardino & Dalrymple 1992), koalas (Canfield 1991), wolves (Mech 1989), turkeys (Holbrook & Vaughan 1985), badgers (Davies et al. 1987), and other vertebrates (Lalo 1987, Putman 1997), practically nothing is known about the impact that roadways have on invertebrates (Seibert & Conover 1991). In fact , a recent book on butterfly conservation (New 1997) makes no mention of the subject, and roads are mentioned only briefly in two other recent books on insect conservation, as barriers to butterfly movement (Samways 1994, Pullin 1995). Samways (1994) states "Roads are line corridors that can cause high mortality where traffic volume is high. In 1989, 100 m of Tennessee roadside was a graveyard for over 120 traffickilled butterflies" (Samways 1994:117). Other than these sources, the best information on road mortality of butterflies is by Munguira and Thomas (1992) in England. They found that roads were not a serious barrier to butterfly movement, but that vehicles killed up to 7% of adult butterflies from some populations. Illinois has 2050 miles of interstate, 276 miles of toll road, 14,892 miles of highway, and 120,782 miles of county, municipal, and other roads (Illinois Department of Transportation pers. com. 1998). Although this comprises the third largest state highway system in the United States, nothing is known about the magnitude of lepidopteran mortality along Illinois roadways. This study was conducted to investigate the magnitude of roadway mortality of Lepidoptera in central Illinois. To quantify the number and kinds of Lepidoptera killed, dead adult Lepidoptera were collected, counted, and identified from along 13 roadside transects in the vicinity of ChampaigniU rbana 400 2'N, 88°17W, Champaign Co., Illinois, with collections occurring weekly on each transect for six weeks. This study is the first to document systematically the magnitude of roadway mortality of Lepidoptera anywhere in the United States. MATERIALS AND METHODS Eleven roadside transects were chosen at the outset of this study to represent different rates of traffic and different roadway types (Table I ). Traffic rates ranged from 0-26,000 vehicles per 24 hour period (Illinois Department of Transportation (lOOT) pers. com. ). To facilitate comparisons, traffic rates above 500 vehicles per day were rounded to the nearest 100. Each roadside transect fit into one of the following general plant community types: remnant prairie, agriculture, or woodland. Roadways were classified into four types: 64 JOURNAL OF THE LEPIDOPTERISTS' SOCIETY TABLE 1. Transect Characteristics. Roadside Speed Vehicles/24 Transect Transect Roadway community Transect limit hours ** numher nmne type type * length (m) (mph ) (#Vehicles ) I Highway 150 #1 Highway Prairie/Ag 150 55 1,000 2 Highway 150 #2 Highway Prairie/Ag 250 5.5 1,000 3 Highway ISO #.3 Highway Prairie/Ag 180 55 1,000 4 Cunningham #1 Divided Ilighway Ag 180 5.5 ]3,500 5 Cunningham #2 Divided Highway Ag 250 55 13,500 6 Cunningham #3 Divided Highway Ag 180 5.5 13,500 7 1-74 Interstate Ag 200 65 26,000 8 M-,57 Interstate Ag 200 6.5 19,700 9 Trelease \"Ioods Country Road Old Field/Woodland 160 35 150 10 Brownfield Woods Country Road Woodland 200 0-35*** 1000 II Airport Grass Control N/A (Mowed Airfield) Mowed Grass 100 N/A 0 12 Country Road Control Country Road Ag 200 0-45*** 150 13 Highway 150 Paved Country Road Prairie/Ag 600 45 50 * Ag ~ Agriculture. **The number of vehicles per day (24 hour period) was obtained through the Illinois Division of Highways . *** The range of speeds given are those typically observed for vehicles accelerating from the stop signs at the beginning of these transects. paved country road, highway, divided highway, and interstate. One additional transect was chosen to control for the effects of roadside mowing. A second additional transect was chosen to quantify the baseline mortality in mowed grass, a habitat rese mbling the mowed roadsides of all transects . Two transects were abandoned one week into the study due to road construction. These transects were replaced with two new transects, and two interstate transects were also added, for a total of 13 transects (Table 1). The nine original transects were first sampled on 25 August 1998. The two replacement transects were first collected on 2 September 1998, and the two interstate transects were first collected on 9 Scptember 1998. We collected along transects weekly until 19 October] 998. Over the two weeks after that date, no dead Lepidoptera were found and the sampling was terminated. Once every seven days each transect was walked and road-killed Lepidoptera were collected into a plastic bag. All 1.3 transects were collected ovcr a two-day period each week. Transects with a median were searched on all sides of the road, including both sides of the median. As a safety precaution, the medians of interstate transects were not sampled. Both sides of the road were walked against the flow of traffic. Lepidoptera were collected that lay dead within two meters of the edge of the road. This usually included the shoulder and about one meter of "ditch". Lepidoptera lying on the roadway itself were not collected, but fewer than 1 0 such individuals were noticed throughout the entire study. Transect length varied from 100 to 610 m, with a mean of 219 m and standard deviation of 121 m (Table I ). A total of 2850 m of transect was sampled during weeks when all 13 transects were sampled. The Thomasboro, Illinois office of the Illinois Division of Highways repOlted that transect locations received similar roadside maintenance, but on different days. Typical maintenance included mowing and trash removal. After collection, the Lepidoptera were sorted by species or speci.es-group and were counted. They were sorted into six taxonomic groups: 1. Hesperiidae, mostly Epargyreus clarus (silverspotted skipper). 2. Lycaenidae, which were not identified below the family level. 3. Moths , mostly Arctiidae and Noctuidae, grouped together because too few individuals from most families were collected to warrant separate treatment. 4. Nymphalidae, including Danaus plexippus (monarch ), Euptoieta claudia (variegated fritillary), Libytlieana carineta (American snout butterfly) , Limenitis archippus (viceroy), Limenitis arthemis astyanax (red-spotted purple), Phyciodes tharos (pearl crescent) , Polygonia interrogationis (question mark), Junonia coenia (common buckeye) , and Vanessa cardui (painted lady). 5. Papilionidae, represented in the roadside transects only by Papillo polyxenes (black swallowtail). 6. Pieridae, including Colias eurytheme (yellow sulfur) and Coli as philodice (clouded sulfur) , which due to hybridization and difficulty in separation into species were grouped together and called the C. eurytheme/C. philodice hybrid complex. Pieridae also included Eurema lisa (little yellow) and Pieris rapae (cabbage whitc).