We conducted this study from January 2009 through January 2015 at Marine Corps Air Station (MCAS) Cherry Point, Craven County, North Carolina (lat 34°54′N, long 76°52′W). MCAS Cherry Point is a U.S. Department of Defense aviation facility located on the south side of the Neuse River basin adjacent to Havelock, NC approximately 80 km inland from the Atlantic Ocean. Mean annual precipitation at the study area is 1,300 mm yr− 1 with 60% typically falling as rain during April – September (Goodwin 1989). The average daily temperature during summer is 26.1 ºC and 8.3 ºC during winter. Soils on the study area consisted of Norfolk loamy fine sands (very strongly acidic, well-drained, moderate permeability), Bragg soils altered by construction methods (extremely acidic, well-drained, moderate permeability), and Rains fine sandy loams (extremely acidic, poorly drained, moderate permeability) (Goodwin 1989).
Grassland habitats on the MCAS Cherry Point airfield are managed in accordance with air safety regulations and mowed during the growing season. These grassland areas consist of a variety of plants, whose origins are likely from post-airfield construction seeding efforts and from the herbaceous layer of Mesic Pine Flatwoods forest communities (Schafale and Weakley 1990) that surround the airfield grasslands. Dominant plants on the study area included tall fescue (Lolium arundinaceum (Schreb.) S.J. Darbyshire), bahiagrass (Paspalum notatum Flueggé), little bluestem (Schizachyrium scoparium (Michx.) Nash), hairy crabgrass (Digitaria sanguinalis (L.) Scop.), trumpet creeper (Campsis radican (L.) Seem. ex Bureau), goldenrod (Solidago sp. L.), poison ivy (Toxicodendron radicans (L.) Kuntze), and Virginia creeper (Parthenocissus quinquefolia (L.) Planch.).
An integrated wildlife damage management program is conducted at MCAS Cherry Point to reduce the risk of collisions between wildlife (e.g., birds, deer) and military aircraft. Birds common to suburban and grassland areas in coastal areas of North Carolina (e.g., American robin [Turdus migratorius], eastern meadowlark [Sturnella magna], European starling [Sturnus vulgaris], laughing gull [Larus atricilla], red-tailed hawk [Buteo jamaicensis]) are commonly observed on the airfield at MCAS Cherry Point. We used USDA ITIS (2015) as the source for plant and animal scientific nomenclature.
Prescribed Burning Applications and Study Plots
In November 2011, we established two sets of paired study plots that each contained a 5.25-ha control (unburned) and a 5.25-ha prescribed burn monitoring plots in the grassland areas of the MCAS Cherry Point airfield (Fig. 1). Two plots were established in a northern area of the airfield grasslands that has never received any biosolids. The other two plots were established in a southern area of the airfield grasslands that has annually received surface-applied lime-stabilized biosolids (e.g., sewage sludge) at a rate of 7.6 Mg ha− 1 on a dry weight basis (using an average of 194 887 L ha− 1 of water as a carrier) annually during a 27-yr period (1989–2015). The two sets of paired plots were approximately 2 km apart and were similar in distance to forested areas, runways, and other landscape characteristics. Plant communities in the plots were dominated by herbaceous vegetation, consisting of a variety of grasses, forbs, and woody plants.
Prescribed burning activities were conducted during January of 2012 and February of 2013 and 2014. We conducted winter burning activities in late afternoon and early evening when wind was low (2–16 kph), relative humidity was rising (25–60%), and air temperatures ranged from 10 to 20°C (Packard and Mutel 1997).
Throughout the growing season (April – October), we measured vegetation height each month during 2012, 2013, and 2014. During each month, 30 sample points were randomly selected in each of the 4 study plots using a random numbers table. At each sample point, we measured the vegetation height by placing a 1-m stick vertically and recording the average height (in cm) of living vegetation surrounding the stick (i.e., within 25 cm of the stick).
Plant communities were further described by randomly establishing and sampling 30 1-m2 herbaceous sampling plots within each study plot during fall of 2011 (1–3 November; prior to prescribed burning), spring of 2012 (23–24 April), fall of 2012 (22–23 October), spring of 2013 (29–30 April), fall of 2013 (13–15 November), spring of 2014 (6–7 May), and fall 2014 (28–29 October). We visually estimated the total vegetative canopy cover (%), bare ground (%), litter (%), and canopy cover (%) of each individual plant species for each herbaceous sampling plot (Bonham 1989). Plant species richness was determined by identifying and counting the total number of different plant species within each herbaceous sampling plot (Bonham 1989). In addition, we determined the relative percent plant community composition of 4 vegetation classes (i.e., grasses, forbs & legumes, vines, and woody plants) by totaling the percent cover of all plants categorized within the vegetation classes within each individual herbaceous sampling plot (Bonham 1989).
Bird surveys (3-min fixed area point-counts; Ralph et al. 1995; Bibby et al. 2000) were conducted each month from January 2013 through January 2015, equally distributed among three different time periods (i.e., morning, mid-day, evening). Bird observations were conducted an average of 5.6 days month−−1 (range = 2–7) starting at randomly chosen plots and times. We observed each study plot from a fixed point within 30 m of the center of the plot for 3 min once during each bird survey. The number of birds observed on the ground or on a plant within the plot, flying and feeding over the plot, or flying over the plot was recorded by species and activity.
White-tailed Deer Surveys
We estimated white-tailed deer (Odocoileus virginianus) use of the unburned and prescribed burn study areas, as well as the entire airfield. During 2009–2014, we conducted a total of 148 white-tailed deer surveys (average of 24.7 surveys yr−−1) associated with the four study plots. White-tailed deer surveys began approximately 30 min after sunset. During each survey, we used a pick-up truck to travel to each of the four study plots. Using a 1 000 000 candle power spotlight, two observers examined each study plot for a 5-min period and counted the total number of white-tailed deer observed within each plot.
In addition to the deer surveys conducted in the study plots, we evaluated historical and concurrent MCAS Cherry Point airfield white-tailed deer surveys (n = 148; 2.1 surveys per month) conducted by USDA Wildlife Services during 2009–2014. The entire white-tailed deer spotlight survey route was approximately 26.2 km in length and encompassed all portions of the airfield.
Vegetation data (mean vegetation height and plant community composition) were normally distributed. We used 2-way analysis of variance (ANOVA) (Zar1996) to compare vegetation height between control and prescribed burn plots as well as between biosolids-treated and those without biosolids for 2012, 2013, and 2014 independently.
We used two-way analysis of covariance (ANCOVA) and Fisher’s Protected LSD tests to compare the plant community characteristics (i.e., total vegetative canopy cover, bare ground, litter, and plant species richness) among the four treatments and six sampling seasons (i.e., spring and fall each year) (Neter et al. 1990, Zar 1996). We used the appropriate pre-treatment (i.e., fall of 2011) plant community characteristics as a covariate.
Two-way ANCOVA and Fisher’s Protected LSD tests were also used to compare the vegetation composition components (i.e., grass, forbs & legumes, vines, and woody plants) among the four treatments and six sampling seasons (i.e., spring and fall each year) (Neter et al. 1990 Zar 1996). We used the appropriate pre-treatment (i.e., fall of 2011) vegetation composition components as a covariate.
We wanted to consider only birds actually associated with (e.g., using) the plots and thus removed birds with the “flying” activity codes from the data prior to analyses. Additionally, we assigned all birds observed using the control or biosolids-treated monitoring plots into foraging guilds using a standard classification (DeGraff et al. 1985). We compared bird use (number of birds per 3-min survey) among the four treatments using repeated measures ANOVA and Fisher’s Protected LSD tests (Neter et al. 1990, Zar 1996). We compared the proportion of birds within foraging guilds using the four treatment plots using G-tests for goodness-of-fit tests (Zar 1996).
We compared white-tailed deer use (number of deer observed per plot) among the four treatments and before and after burning implementation using two-way repeated measures ANOVA and Fisher’s Protected LSD tests (Neter et al. 1990, Zar 1996). In addition, we compared white-tailed deer relative abundance (total number of deer observed per airfield survey) on the MCAS Cherry Point airfield during 2009─2011 to the relative abundance of deer on the airfield during 2012─2014 using repeated measures ANOVA and Fisher’s Protected LSD tests (Zar 1996). We considered differences significant at P ≤ 0.05 and conducted all analyses using SAS statistical software version 9.1 (SAS Institute, Cary, NC). Data are presented as mean ± 1 standard error (SE).