Effects of white-tailed deer (Odocoileus virginianus Zimm.) herbivory in restored forest and savanna plant communities.
|Abstract:||Despite the widespread interest in plant community restoration, few studies have assessed white-tailed deer (Odocoileus virginianus Zimm.) herbivory on herbaceous species and even fewer studies have focused on deer herbivory in restored plant communities. During 2007-2009, we investigated the effect of deer density and associated deer browse on two restored forest and three restored savanna plant communities in Lake County, Illinois. We used 300 small (1.4 m diameter x 1.6 m height) exclosed plots and 1-[m.sup.2] unexclosed plots to compare the effects of deer herbivory on forbs. We quantified and compared percent nonherbaceous ground cover, species diversity, species evenness, and floristic quality between exclosed plots and unexclosed plots, as well as among preserves within each plant community type. Species diversity and floristic quality of forbs may be maximized at a deer density between 6-22 deer [km.sup.2] in restored forest communities in northeastern Illinois. Floristic quality was higher in exclosed plots compared to unexclosed plots at all savanna sites. In both plant communities, species evenness may have increased with higher deer density due to an increase in non-preferred plants and non-native species invading locations where preferred native forbs were chronically consumed. Our results highlight the importance of assessing the species diversity, evenness, and floristic quality of target plant communities to determine the impact of deer herbivory at varying deer densities.|
Plants (Environmental aspects)
White-tailed deer (Physiological aspects)
White-tailed deer (Environmental aspects)
Urbanek, Rachael E.
Nielsen, Clayton K.
Glowacki, Gary A.
Preuss, Timothy S.
|Publication:||Name: The American Midland Naturalist Publisher: University of Notre Dame, Department of Biological Sciences Audience: Academic Format: Magazine/Journal Subject: Biological sciences; Earth sciences Copyright: COPYRIGHT 2012 University of Notre Dame, Department of Biological Sciences ISSN: 0003-0031|
|Issue:||Date: April, 2012 Source Volume: 167 Source Issue: 2|
|Topic:||Event Code: 310 Science & research|
|Geographic:||Geographic Scope: United States Geographic Code: 1USA United States|
White-tailed deer (Odocoileus virginianus Zimm.) are keystone herbivores known to directly impact plant communities via browsing. Rooney and Waller (2003) noted that overbrowsing by deer in mixed coniferous-deciduous forests causes ferns, grasses, sedges, and rushes to become the dominant herbaceous understory. Intense selective browsing on forb species by deer may cause local extirpation of some herbaceous plant species (Strole and Anderson, 1992) because a single deer can consume all above-ground tissue, including reproductive flowers, in a single bite of some herbaceous plants (Russell et al., 9001). Although a single ungulate can completely defoliate a forb, plants often survive by physiologically reducing growth and reproduction in the following growing season which deters deer from future herbivory (Edwards, 1985; Whigham, 1990; Primack et al., 1994; Rooney and Waller, 9003). Overbrowsing may lead to local extirpation if this occurs on a yearly basis to a preferred species. Conversely, deer may also aid in seed dispersal of select forb species through their feces, which may negate extirpation (Vellend et al., 2003; Myers et al., 2004).
Many natural resource management agencies aim to restore plant communities to the conditions that existed prior to human development (Dobson et al., 1997; Cole, 2002; Suding et al., 2004). However, plant community restoration efforts may be hindered by abundant white-tailed deer populations. High deer densities reduce the diversity and height of woody plant regeneration through herbivory (Marquis, 1974, 1981; Tilghman, 1989; Ruzicka et al., 2009). Distance from woodland edge and edge structure may influence the location of where deer browse, which ultimately affects the composition and density of tree seedlings (Cadenassa and Pickett, 2000; Ruzicka et al., 2009). Actively growing shoots and leaves of young woody species are preferred by deer, and overbrowsing can lead to the plant growing wider rather than taller (Gill and Beardall, 2001). Tree seedlings may become readily established in forest regenerations, but the number of seedlings that develop into saplings is strongly negatively correlated to deer density (Tilghman, 1989). As a result of deer reduction, hardwood regenerations have been reported to increase eight-fold in a 5 y period (Behrend et al., 1970).
Despite the widespread interest in plant community restoration, few studies have specifically assessed deer herbivory in restored plant communities. Several studies have examined the effects of deer herbivory on tree regeneration (Marquis, 1974, 1981; Tilghman, 1989; Cadenassa and Pickett, 2000; Tremblay et al., 2006), but to our knowledge, few studies involved restored herbaceous plant communities. Augustine and Frelich (1998) reported browsing effects on restored forbs, noting an increase in the proportion of smaller size classes and a general inhibition of growth in transplanted trillium (Trillium sp. Michx.) communities. These trillium communities also experienced a large reduction in flowering rates at high deer densities. Englund and Meyer (1986) noted that deer herbivory did not affect prairie forb seedling survival; however, deer herbivory did reduce stem height and number of reproductive flowers on browsed plants. Conversely, Ruhren, and Handel (2003) found a decrease in survival and flowering of transplant forest forbs growing in exclosed versus unexclosed plots exposed to deer herbivory. In the most recent study, Anderson et al. (2007) found that intense deer browsing in combination with the initial number of species sowed can limit the quality of a prairie restoration.
With little information available in the literature, resource managers are uncertain about whether deer herbivory influences the success of plant community restorations. Although restoration managers often draw from literature on the effects of deer herbivory on natural communities, managers would benefit from an assessment of deer herbivory impacts on lands that are actively being restored. We conducted our study in suburban forest preserves managed by the Lake County Forest Preserve District (hereafter, District), Illinois. Similar to many other land management agencies, the District aims to restore high quality plant communities in forests and savannas. Our objective was to quantify the effects of deer herbivory on restored forest and savanna plant communities to provide further insight to natural resource managers aiming to restore native plant communities.
Lake County, Illinois, is located approximately 40 km northwest of downtown Chicago, has >700,000 inhabitants, and is <1200 [km.sup.2] in area (USCB, 2009). From 1971 to 2000, mean temperature was 8.5 C ranging from -32.8-38.9 C (NOAA, 2004). Total annual precipitation averaged 86.6 cm (NOAA, 2004). Predominant soil types are silt loam and silt clay (Paschke and Alexander, 1970).
The District, created in 1958, consists of >10,800 ha in 57 preserves that provide a combination of natural, recreational, educational, and cultural resources for county residents and tourists. The District identifies four separate plant communities within its preserve system as characterized by the Chicago Wilderness Terrestrial Classification System with forests containing >80% canopy cover and savannas 10-50% canopy cover (CRBC, 1999).
Restored forb communities investigated in this study were defined as 10-ha parcels of land containing native flora present before European settlement. These parcels of land were located within five District preserves that we chose based on 2007-2009 overwintering deer density estimates (Urbanek et al., 2012a) and plant community restoration type. We investigated the effects of deer herbivory on two preserves that had forest restoration sites (high density forest = 22 deer/[km.sup.2]; low density forest = 6 deer/[km.sup.2]) located within there and three preserves that contained savanna restoration sites (high density savanna = 30 deer/[km.sup.2]; medium density savanna = 21 deer/[km.sup.2]; low density savanna = 13 deer/[km.sup.2]). Due to the nature of this suburban landscape (i.e., preserves in this study were fragments of natural areas isolated in an urban matrix), deer were known to be year-round residents on these preserves. Deer management via professional sharpshooting occurred only on the low density forest preserve; all other preserves had no deer management or public harvest seasons. Deer densities were obtained using pellet-based distance sampling and program DISTANCE (Thomas et al., 2010) for each preserve in Mar.-Apr. 2008 and 2009 (Urbanek et al., 2012a). Deer densities did not differ among years (Urbanek et al., 2012a), thus we used the average density over the course of the study for each preserve.
Forest restorations were mesic upland forests that included mostly oak (Quercus spp.), black raspberry (Rubus occidentialis L.), European buckthorn (Rhamnus cathartica L.), honeysuckle (Lonicera spp.), and sugar maple (Acer saccharum Marsh.). Restoration on both forest sites began in the mid-1990s which has included the continual removal of nonnative woody species and yearly prescribed burns. No seeding or tree planting has occurred on either of these forest restoration sites. Savanna restorations were mesic savannas with a mixture of mixed prairie forbs interspersed with white oaks (Q. alba L.) and bur oak (Q. macrocarpa Michx.). On all sites, drain tiles have been removed, native plant seeding has occurred, and prescribed burns have been conducted on a yearly basis. Unlike the forest restoration sites which were similar in appearance, the savanna restoration sites appeared to be at different stages of restoration. The low density savanna was an open prairie with 100 oak saplings planted in 2001. Conversely, the high density savanna was more forested and additional restoration practice included the removal of teasel (Dipsacus lacinatus L.). Teasel was also removed on the medium density savanna, and the site resembled a tallgrass savanna comprised of an extensive open area interspersed with large, mature oak trees that included plants found in both prairies and forests.
Thirty small (1.4 m diameter x 1.6 m height) exclosures were spaced in each site in a grid design approximately 60 m from any neighboring exclosure and 30 m from any road, path, railroad, or habitat edge. An unexclosed 1-[m.sup.2] plot was located 30 m away from each exclosure at an azimuth of 225[degrees] to accommodate all unexclosed plots within the designated restoration area. Unexclosed plots were also [greater than or equal to]30 m from any other exclosed plot, road, path, railroad, or habitat edge. This spacing design allowed us to examine the majority of area being restored. Most studies of deer herbivory have used a few large exclosures adjacent to unexclosed sites (Marquis, 1974, 1981; Casey and Hein, 1983; Anderson et al., 2005). With that study design, bias exists because plants are not naturally randomly distributed and large exclosures may not contain browsed species from a different portion of the study site. Thus, smaller, but more abundant, exclosures are preferable to increase sample size and reduce bias. We did not collect pre-treatment data in this study because we believe our plot placement in this grid design ensured that there were no pre-existing differences between the exclosed and unexclosed plots since they were distributed over the entire restoration site.
We constructed deer exclosures during Jun.-Aug. 2007 with 5.4 x 1.6-m rectangular pieces of fixed-knot 12.5-gauge (ga) wire STAY-TUFF[R] horse fencing (Stay-Tuff Fence Manufacturing, Inc., New Braunfels, Texas, USA) by bending the pieces to form circular exclosures with approximately 1.4-m diameters. This size allowed for a 1-[m.sup.2] quadrat to fit within the exclosure with a 0.2-m buffer to allow deer browsing in the periphery of the exclosure. The horizontal spacing between stay wires gradually diminished from 18 cm at the bottom to 8 cm at the top and the vertical distance between line wires remained constant at 15 cm. This design allowed for smaller browsing animals [e.g., eastern cottontails (Sylvilagus floridanus Allen), woodchucks (Marmota monax L.)] to enter the exclosure in the larger spacing at ground level while deterring deer from browsing at the top. Exclosures were tied with 16-ga wire to three 1.2-m posts staked at angles from each other for stability.
We collected plant measurements within the exclosed plots and unexclosed plots during 1-30 Jun. 2008-2009. We used similar deer browse measurements as found in the literature (Anderson, 1990, 1994; Augustine and Frelich, 1998; Webster and Parker, 2000; Frankland and Nelson, 2003). All herbaceous plants found within the unexclosed plots were identified and number of individual plants of each species was tallied. The number of ramets (hereafter, plants) browsed by deer was quantified per species. We only included browsed plants that we were 100% confident were browsed by deer and not by any other animal. Plants browsed by deer were identified by jagged edges as opposed to rodent browse which tends to be angled and clean-cut (Anderson, 1969; Hygnstrom et al,, 1994). We identified and quantified the number of individual plants of all species in exclosed plots. In all plots, we estimated percent cover of non-vegetated soil, grass, and sedge.
We used the same methods to analyze data for each plant community type. We used a chi-square test ([alpha] = 0.05 throughout) to determine whether the number of browsed versus unbrowsed species differed among sites. We used the Shannon Index (Shannon and Weaver, 1949) and Simpson's Dominance Index (Simpson, 1949) as measures of diversity and evenness in each exclosed and unexclosed plot. Differences in diversity and evenness among exclosed and unexclosed plots and among preserves were determined using a Model I 2-way ANOVA. We analyzed these indices three ways: (1) for all forbs combined, (2) for native species only, and (3) for non-native species only. We also used a Model I 2-way ANOVA to compare the cover of non-vegetated soil, grass, and sedge between treatments and among preserves.
The Floristic Quality Index (Swink and Wilhelm, 1994) was also calculated for all exclosed and unexclosed plots. This index is based on native species that have been qualified with "coefficient of conservation" (C-value) values ranging from 0-10 (Swink and Wilhelm, 1994). Plants that are remnant of a natural community and appear rare are given a value of 10, whereas those species that are highly adaptable to disturbance and are common have values of 0 or 1 (Swink and Wilhelm, 1994). An area's total floristic quality is determined by the equation: FQI = [bar.C][square root of N] where FQI is the floristic quality index value, C is the mean C-value and N is the total number of native species found in the area (Swink and Wilhelm, 1994). A Model I 2-way ANOVA was used to determine the floristic quality differences between unexclosed and exclosed plots and among preserves.
[FIGURE 1 OMITTED]
In the forest sites, we identified 60 forb species at the low density forest and 83 forb species at the high density forest (Fig. 1). There was no difference in the number of browsed versus unbrowsed plants between forest sites ([X.sub.1.sup.2] = 0.10, P > 0.10). Deer at the low density forest site browsed all native species, whereas we found deer browse at the high density forest included four non-native species. Of the savanna sites, the high density savanna preserve had the most species (Fig. 1), followed by the medium and low density savanna preserves (n = 112; 95; 56, respectively). The number of browsed versus unbrowsed species differed when we compared all three savanna sites ([X.sub.2.sup.2] = 8.17, 0.010 < P < 0.05). Upon further analyses, the number of browsed versus unbrowsed species did not differ between the high and medium density savanna preserves ([X.sub.1.sup.2] = 0.05, P > 0.10), thus the difference in the previous analyses was driven by the number of browsed versus unbrowsed species in the low density savanna. Therefore by this measure, browsing was heaviest in the high and medium density savannas. The medium density savanna had the most non-native species browsed (n = 7); both the low and high density savanna had four non-native species browsed each.
In the forest sites, both measures of species diversity (Shannon Diversity Index and Simpson's Dominance Index) indicated that the high density forest had a greater combined species diversity (includes non-native and native species), native species diversity, and invasive species diversity than the low density forest (Site: [F.sub.1,165-236] = 5.48-44.33, P [less than or equal to] 0.01; Fig. 2). For evenness index analyses, an effect was observed when we analyzed all species combined and for native species using the Shannon Diversity Index (Site: [F.sub.1,235-236] = 9.7314.58, P [less than or equal to] 0.01) but not for the Simpson's Dominance Index (Site: [F.sub.1,236] = 0.27-0.61, 0.44 [less than or equal to] P [less than or equal to] 0.61). Conversely, we observed an effect for non-native species using the Simpson's Dominance Index (Site: [F.sub.l,116] = 16.33, P < 0.01) but not for the Shannon Diversity Index (Site: [F.sub.1,166] = 2.49, P = 0.12). In general, the high density forest had slightly higher evenness values than the low density forest preserve for all evenness analyses (Fig. 2). Higher percentages of grass, sedge, and non-vegetated soil cover were found at the low density forest than at the high density forest (Site: [F.sub.1,114-236] = 2.36-16.90, 0.01 [less than or equal to] P < 0.05; Fig. 3). Floristic quality was higher at the high density forest compared to the low density forest (Site: [F.sub.1,236] = 9.31, P < 0.01; Fig. 4).
[FIGURE 2 OMITTED]
In the savanna sites, both diversity indices indicated the lowest combined species diversity was found at the medium density savanna and the highest native species diversity was found at the high density savanna (Site: [F.sub.2,353-354] = 35.00-46.07, P < 0.01; Fig. 5). The highest diversity of non-native species was detected at the low density savanna, followed by the medium and then the high density savannas (Site: [F.sub.2,268-329] = 42.72-66.85, P < 0.01; Fig. 5). Using the Shannon Diversity Index ([F.sub.2,347-353] = 5.36-7.18, P < 0.01), we detected an effect on evenness for all species combined and native species only; however, we did not observe this effect using the Simpson's Dominance Index ([F.sub.2,353-354] = 1.22-1.60, 0.20 [less than or equal to] P [less than or equal to] 0.30). Using the Shannon Diversity Index, the medium density savanna had the lowest evenness when all species were combined and the high density savanna had the highest evenness for native species only (Fig. 5). Both indices depicted differences among sites for non-native species and the medium density savanna appeared to have the highest evenness (Site: [F.sub.2,268-329] = 8.30-43.95, P [less than or equal to] 0.01; Fig. 5). We detected the highest percent grass cover at the medium density savanna, followed by the low and then the high density savanna (Site: [F.sub.2,354] = 89.64, P [less than or equal to] 0.01; Fig. 3). The low density savanna had the lowest non-vegetated soil percent cover (Site: [F.sub.2,354] = 89.64, P [less than or equal to] 0.01; Fig. 3). There was no difference in percent sedge cover among savanna sites (Site: [F.sub.2,79] = 0.09, P = 0.91; Fig. 3). The highest floristic quality was found at the high density savanna, followed by the medium density savanna, and then the low density savanna (Site: [F.sub.2,353] = 74.54, P < 0.01; Fig. 4).
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
In the forest sites, unexclosed plots at the low density forest had greater combined species diversity and native species diversity than in exclosures (Interaction: [F.sub.1,235-236] = 5.48-9.40, 0.01 [less than or equal to] P [less than or equal to] 0.02; Fig. 2). However, exclosed plots at the high density forest had greater combined species diversity and native species diversity than in the unexclosed plots (Fig. 2). No exclosure or interaction effects were observed for non-native species in the forest sites ([F.sub.1,165-166] = 0.03-2.85, 0.09 [less than or equal to] P [less than or equal to] 0.86). Thus, differences in native species are the main variable driving differences in diversity between treatments. No exclosure or interaction effects were observed in any evenness analysis in the forest sites ([F1.sub.88-236] = 0.01-2.73, 0.10 [less than or equal to] P [less than or equal to] 0.92; Fig. 2). A higher percentage of grass cover (Exclosure: [F.sub.1,173] = 4.49, P = 0.04) was found in the unexclosed forest plots, whereas a higher percentage of sedge cover (Exclosure: [F.sub.1,114] = 15.65, P < 0.01) was observed in the exclosed forest plots (Fig. 3). No exclosure effect was detected for non-vegetated soil (Exclosure: [F.sub.1,236] = 3.04, P = 0.08) in forested sites. No interaction effects were found for percent grass, sedge, and non-vegetated soil cover (Interaction: [F.sub.1,236] = 0.10-1.15, 0.28 [less than or equal to] P [less than or equal to] 0.75). We found higher floristic quality in the unexclosed plots at the low density forest compared to exclosed plots, whereas the floristic quality in exclosed plots at the high density forest was higher than in the unexclosed plots (Interaction: [F.sub.1,236] = 9.31-20.5, P < 0.01; Fig. 4).
[FIGURE 5 OMITTED]
In the savanna sites, the Simpson Dominance Index (Interaction: [F.sub.2,353] = 3.47, P = 0.03) was significant for the combined species diversity; however, there was no clear trend for the interaction effect (Fig. 5). The Shannon Diversity Index (Interaction: [F.sub.2,354] = 1.03, P = 0.36) was non-significant for combined species diversity. No exclosure effect or interaction effects on species diversity was observed for native species or non-native species in the savanna sites ([F.sub.2,268-353] = 0.01-1.16, 0.32 [less than or equal to] P [less than or equal to] 0.93). Unexclosed savanna plots had higher evenness values for combined species and native species than exclosed plots in the low and medium density savannas (Interaction: [F.sub.2,347-354] = 2.90-5.39, 0.01 < P < 0.05; Fig. 5). Conversely, exclosed plots had higher combined species evenness and native species evenness at the high density savanna site (Fig. 5). The Shannon Diversity Index (Interaction: [F.sub.2,268] = 0.9, P = 0.41) and Simpson's Dominance Index (Interaction: [F.sub.2,329] = 3.21, P = 0.04) did not agree for non-native species evenness. According to the Simpson's Dominance Index, nonnative species evenness was higher in exclosed plots at medium and high density savannas, whereas non-native species evenness was higher in unexclosed plots at low density savannas (Fig. 5). A higher percentage of grass cover was observed in unexclosed plots compared to exclosed plots at the medium and low density savannas (Interaction: [F.sub.2,354] = 4.27, P = 0.01; Fig. 3). Conversely, a higher percentage of grass cover was observed in exclosed plots compared to unexclosed plots at the high density savanna (Fig. 3). Similarly, a higher percentage of non-vegetated soil was observed in exclosed plots compared to unexclosed plots at the low and medium density savannas, whereas a higher percentage of non-vegetated soil was found in the unexclosed plots at the high density savanna (Interaction: [F.sub.2,339] = 11.28, P < 0.01; Fig. 3). No exclosure or interaction effects were found for percent sedge cover (F2,79 = 0.41, 0.52 [less than or equal to] P [less than or equal to] 0.66). Exclosed plots had higher floristic quality than unexclosed plots (Exclosure, [F.sub.1-2,353] = 20.657, P [less than or equal to] 0.01; Fig. 4).
Local plant studies are needed to assess deer herbivory impacts and inform management decisions (Strole and Anderson, 1992). Often these studies have limited focus and only pertain to the deer density on the study area investigated. To our knowledge, we are the first to investigate the effect of varying browse pressure on forbs via multiple deer densities on forest and savanna plant restorations. Other studies that examined the effects of deer herbivory on restorations did not compare areas that differed in deer density but rather just compared the effects on restored plants in the absence of deer herbivory versus exposed restored forbs (Englund and Meyer, 1986; Augustine and Frelich, 1998; Anderson et al., 2001; Ruhren and Handel, 2003). Our study contributes to the sparse knowledge specific to restoration management available to aid in the restoration of forest and savanna plant communities given abundant deer populations are often present on natural areas.
No plant community measurements were taken prior to the first field season for either plant community. Thus, although the forested preserves appeared similar in physical structure (i. e., with apparently similar species composition, vertical, and horizontal structure), we cannot be positive of this claim. As aforementioned in the study site descriptions, the savanna sites were in different stages of restoration with obvious structural differences. Hence, for both plant communities we focus our discussion on site and treatment (i.e., exclosure) differences and discuss how these differences may be a reflection of different chronic deer densities.
White-tailed deer are concentrate selectors that browse on a large variety of plants to identify their most preferred forage (Hofmann, 1989). Furthermore, browsing pressure on forbs is often positively correlated with deer density (Augustine and Jordan, 1998; Frankland and Nelson, 2003). Deer are also primarily forest-edge inhabitants that use forested areas for cover and the majority of their foraging (Russell et al., 2001). Although the high density forest had more deer and higher species richness, it is unsurprising that we did not find a difference between the number of browsed versus unbrowsed species among forest sites because deer are known to select preferred species such as trillium and leave many other species untouched (Anderson, 1994; Augustine and Frelich, 1998). However, we did find a greater number of browsed versus unbrowsed species in the savanna restorations with deer densities [greater than or equal to] 21 deer/[km.sup.2] compared to the site with a deer density of 13 deer/[km.sup.2]. The medium and high density savannas may have provided superior cover (i.e., trees) and preferred forest forbs for deer to remain in these areas to forage, as indicated by higher species richness values. Deer at the low density forest may have selected for preferred native forbs only, therefore, limiting the number of total browsed species. Alternatively, deer at the low density savanna preserve may have browsed in the woodlands surrounding this savanna restoration and not have browsed extensively on the studied plant community itself.
Overbrowsing by deer can cause forbs to be replaced with roughage material such as grasses and sedges (Rooney and Waller, 2003). In the absence of deer browse, exclosed plots had less grass cover than the unexclosed plots in the forested sites. Exclosed plots in the savannas also exhibited less grass cover compared to unexclosed plots at densities [less than or equal to] 21 deer/[km.sup.2]. These results suggest these deer densities and subsequent browse are negatively affecting these plant communities. However, sedge cover was higher in exclosed plots compared to unexclosed plots at the forested sites which is the opposite of what we would expect if high density deer browsing was having a negative effect on this plant community. Similarly, we would have also expected to find higher percentages of grass cover, sedge cover, and non-vegetated soil at the high density forest and increasing percentages of these cover types as deer increased among savanna sites. Instead, we recorded the opposite trend on the forest sites and the highest amounts of grass cover at the medium density savanna. It is widely known that chronically high deer browsing may leave legacy effects (Russell et al., 2001; Banta et al., 2005; Royo et al., 2010b) and 2 y of excluding deer may not be enough time to allow all ground cover composition to recover within the exclosures to exhibit significant differences compared to the unexclosed plots. Thus, the presence of these abundant roughage plants in the exclosed plots in both plant communities may be a legacy of chronic high deer densities. In addition, it is important to note that deer culling began at the low density forest preserve in 1992 when deer density was 44/[km.sup.2]; the preserve has been managed at the current deer density of 6/[km.sup.2] only since 2008. This short recovery time may be the reason why we observed high amounts of roughage on this preserve. Similarly, Royo et al. (2010b) found no difference in abundance of graminoids following the culling of a deer herd over a 6 y period. Conversely, the differences we observed in the savanna sites can mostly likely be attributed to site differences and not impacts of deer density and associated browse pressure.
Overall diversity in a plant community may increase due to the acceleration of the invasion of nonnative species due to high deer density and subsequent canopy disturbances (Eschtruth and Battles, 2009). Similarly, we found higher diversity at the high density forest for all species combined, native species, and invasive species. We also observed slightly higher species evenness at the high density forest possibly because deer have exhausted the supply of preferred species which has allowed more less-preferred native species and invasive species to invade (i.e., thus increasing diversity). Deer at the high density forest may be more euryphagus than deer at the low density forest due to overbrowsing (and potentially extirpating) their preferred and more palatable species (Webb, 1959; Augustine and Jordan, 1998; Suzuki et al., 2008). Deer diets included nonnative species at the high density forest and these deer browsed more species (n = 34) compared to deer at the low density forest, indicating that the higher browse pressure at this site is depleting the most preferred forbs. The interaction effect between exclosure and site treatments we observed for native species diversity also supports the idea that deer at the high density forest site were eating more native forbs. Additionally, the same interaction effect for floristic quality in the forests indicate that deer herbivory at densities as low as 6 deer/[km.sup.2] may aid in increasing these indices; however, when deer densities reach [greater than or equal to] 22 deer/[km.sup.2], effects of deer herbivory may become negative. This same scenario has been identified in conifer plantations and hardwood forests used by sika deer (Cervus nippon Temm.) in Japan (Suzuki et al., 2008). Royo et al. (2010a) also found that deer densities as low as 4.6-7.7 deer/[km.sup.2] enhanced diversity in deciduous forests compared to areas exclosed from all deer herbivory.
The majority of our results for the savanna sites were indicative of the different stages of restoration the savannas were under. The high diversity of combined and native species and high floristic quality found at the high density savanna was likely observed due to the ecotone of forest and prairie vegetation found there. Likewise, the high diversity of nonnative species and low floristic quality found at the low density savanna was most likely a result of the site being in such an early stage of restoration from agricultural field to savanna. However, the evenness indices may have been affected by deer densities and subsequent browse pressure. We observed the highest evenness values for native species at the high density savanna. Similar to the forest sites, this may mean that as deer density increases in savannas, amplification of browse pressure on preferred forbs may allow less-preferred native species and invasive species to increase (i.e., thus increasing diversity and evenness). In fact, deer diets became more euryphagous as deer density increased in the savanna sites and the plant community at the medium density savanna included more nonnative species than at the low density savanna. We would have also expected to observe more nonnative species browsed at the high density savanna. However, the high density savanna was the most forested which may have encouraged deer to forage more heavily on forest-associated forbs and woody vegetation than deer in the other savanna sites where a wider variety of forbs (i.e., forest forbs and prairie forbs) was available. Thus, the structure of this preserve may have provided deer with a wider variety of native forest-related species to browse compared to the other savanna sites. This browse pressure on preferred native forest forbs may also have permitted non-native species to invade and persist, thus providing a high non-native evenness value. Similarly, deer foraging within the low density savanna may have been selecting only native preferred forbs amongst the great diversity of non-native species available there. This scenario would also result in higher evenness values of nonnative species.
We found that floristic quality was higher in exclosed plots than unexclosed plots among all sites except the low density forest site (6 deer/[km.sup.2]), suggesting that deer densities >13 deer/[km.sup.2] (i.e., minimum deer density at all other sites in this study) may negatively affect this metric. Asnani et al. (2006) also found the same results in a suburban Ohio woodland preserve. The exclusion of deer browse had no apparent effects on diversity and evenness on our savanna sites and all differences in index values via significant interactions were attributed to site differences and/or deer density rather than exclosure treatment differences. Similarly, Webster et al. (2005) also found no difference in diversity and evenness values between exclosed and unexclosed plots in areas of chronic high deer densities.
The task of monitoring and assessing the effects of deer herbivory on plant communities is complex. Environmental factors such as soil, canopy closure, and soil nutrient quality all play keys roles in plant survival and growth (Maschinski and Whitham, 1989) which inevitably differs among study sites. These abiotic interactions and changes in deer density and forb preference make straightforward results from herbivory studies rarely, if ever, truly attained. In this study, we found some evidence that deer are negatively affecting forest and savanna plant communities at densities [greater than or equal to] 21 deer/[km.sup.2]. Specifically, our evidence suggests that deer densities [greater than or equal to] 21 deer/[km.sup.2] may reduce native species diversity while increasing nonnative species diversity. In turn, this potentially created a higher evenness of plants at high deer densities due to the proliferation of non-preferred native and non-native species. Additionally, deer browse at these densities negatively affected floristic quality as compared to areas where deer were restricted from browsing. We also found a higher percentage of grass cover in areas exposed to deer browse in both plant community sites which is another known effect of heavy deer browse (Rooney and Waller, 2003). Other results, especially within the savanna plant community, are less clear due to different stages of restoration confounding the treatment and deer density effects. Similarly, legacy effects of chronic deer browse may have shadowed the effects of deer browse on sedge and non-vegetated soil cover in both plant communities. These types of intrusions in deer herbivory studies complicate the decision-making process regarding deer management for managers.
Herein, we provide recommendations to improve future investigations and monitoring efforts of deer herbivory based on the obstacles we encountered in our study. Future studies should incorporate areas containing more than three different deer densities for each plant community to create a more robust depiction of the effect of deer density and subsequent browse on herbaceous communities. An increase in the number of sites would allow biologists to determine which deer densities are optimal for herbaceous growth in their plant communities using regression techniques. Regression analyses may also allow managers to tease out whether differences in plant community metrics are attributable to deer density and subsequent browse, inherent differences in community structure among sites, or a combination of these variables.
We also suggest managers routinely monitor the species diversity, evenness, and floristic quality of their target plant communities to assess the impact of deer populations in their area over time (Anderson et al., 2007; Waller et al., 2009). These metrics will enable managers to compare multiple plant communities to identify the appropriate deer density in their area that will promote restoration efforts without focusing on specific indicator species. In addition, trends of these plant metrics can be monitored to assess the effectiveness of managing deer herds within individual plant communities. We recognize that managers may have heavy work schedules that may limit their ability to collect these data. To alleviate some of the burden, we suggest using browse transects rather than exclosures to assess deer herbivory when deer need not be restricted from specific areas or plants (Urbanek et al., 2012b). This type of deer herbivory monitoring coupled with collecting plant community metrics will result in reduced costs, time, and potentially larger areas surveyed (Urbanek et al., 2011). However, if managers do have the resources to continue monitoring these metrics using exclosures. It may help to determine the length of time required for restoration goals to be met.
Acknowledgments.--We thank J. Anderson and the Lake County Forest Preserve District for funding this project. We thank S. Johnson, W. Mitchell, Z. Robinson, K. Rice, M. Carrillo, J. Haus, T. Rounsville, Jr., and E. Madson for their help in the field. Much gratitude is given to K. Klick and consultants from Witness Tree, INC. for assistance in identifying unknown species. We appreciate the Youth Conservation Corps for constructing exclosures. Two anonymous reviewers and H. Rowe improved a previous draft of this manuscript. Lastly, we thank J. Zaczek, the College of Agricultural Sciences, the Departments of Zoology and Forestry, the Graduate School and the Cooperative Wildlife Research Laboratory at Southern Illinois University Carbondale.
ANDERSON, R. C. 1990. Second interim report on the project to develop methodology to determine the influence of deer browsing on forested natural areas in the Lake County Forest Preserve District. Lake County Forest Preserve District, Libertyville, Illinois, USA.
--. 1994. Height of white-flowered trillium (Trillium grandiflorum) as an index of deer browsing intensity. Ecol. App., 4:104-109.
--, E. A. CORBETT, M. R. ANDERSON, G. A. CORBETT AND T. M. KELLEY. 2001. High white-tailed deer density has negative impact on tallgrass prairie forbs. J. Torrey Bot. Soc., 128:381-392.
--, B. DORICK AND T. CRISPINO. 2007. Impacts of initial species richness and deer browsing on the quality of restored prairie in central Illinois. Ecol. Rest., 25:99-102.
--, D. NELSON, M. R. ANDERSON AND M. R. RICKEY. 2005. White-tailed deer (Odocoileus virginianus Zimmermann) browsing effects on tallgrass prairie forbs: diversity and species abundances. Nat. Areas J., 25:19-25.
ANDERSON, T. E. 1969. Identifying, evaluating, and controlling wildlife damage, p. 497-S20. In: R. H. Giles, Jr. (ed.). Wildlife management techniques. 3rd ed. revised. The Wildlife Society, Washington, D. C., USA.
ASNANI, K. M., R. A. KLIPS AND P. S. CURTIS. 2006. Regeneration of woodland vegetation after deer browsing in Sharon Woods Metro Park, Franklin County, Ohio. Ohio J. Sci., 106:86-92.
AUGUSTINE, D. J. AND L. E. FRELICH. 1998. Effects of white-tailed deer on populations of an understory forb in fragmented deciduous forests. Conserv. Biol., 12:995-1004.
-- AND P. A. JORDAN. 1998. Predictors of white-tailed deer grazing intensity in fragmented deciduous forests. J. Wildl. Manage., 62:1076-1085.
BANTA, J. A., A. A. ROYO, C. KIRSCHBAUM AND W. P. CARSON. 2005. Plant communities growing on boulders in the Allegheny National Forest: evidence for boulders as refugia from deer and as a bioassay of overbrowsing. Nar. Areas. J., 25:10-18.
BEHREND, D. F., G. F. MATTFIELD, W. C. TIERSON AND J. E. WILEY III. 1970. Deer density control for comprehensive forest management. J. Forest., 68:695-700.
CADENASSA, M. L. AND S. T. A. PICKETT. 2000. Linking forage edge structure to edge function: mediation of herbivore damage. J. Ecol., 88:31-44.
CASEY, D. AND D. HEIN. 1983. Effects of heavy browsing on a bird community in a deciduous forest. J. Wildl. Manage., 47:829-836.
CHICAGO REGION BIODIVERSITY COUNCIL (CRBC). 1999. Biodiversity recovery plan. Chicago Region Biodiversity Council, Chicago, Illinois, USA.
COLE, C. A. 2002. The assessment of herbaceous cover in wetlands as an indicator of function. Ecol. Indicat., 2:287-293.
DOBSON, A. P., A. D. BRADSHAW AND A. J. M. BAKER. 1997. Hopes for the future: restoration ecology and conservation biology. Science, 277:515-522.
EDWARDS, J. 1985. Effects of herbivory by moose on flower and fruit production of Aralia nudicaulis. J. Ecol., 73:861-868.
ENGLUND, J. AND W. MEYER. 1986. The impact of deer on 24 species of prairie forbs. Proc. N. Am. Prairie Conf., 9:210-212.
ESCHTRUTH, A. K. AND J. J. BATTLES. 2009. Acceleration of exotic plant invasion in a forested ecosystem by a generalist herbivore. Conserv. Biol., 23:388-399.
FRANKLAND, F. AND T. NELSON. 2003. Impacts of white-tailed deer on spring wildflowers in Illinois, USA. Nat. Areas J., 23:341-348.
GILL, R. M. A. AND V. BEARDALL. 2001. The impact of deer on woodlands: the effects of browsing and seed dispersal on vegetation structure and composition. Forestry, 74:209-218.
HGYSTROM, S. E., R. M. TIMM AND G. E. LARSON. 1994. Prevention and control of wildlife damage. Volume 2. University of Nebraska Cooperative Extension, Lincoln, Nebraska, USA.
HOWANN, T. P. 1989. Evolutionary steps of ecophysiological adaptation and diversification of ruminants: a comparative view of their digestive systems. Oecologia, 78:443-457.
MASCHINSKI, J. AND T. G. WHITHAM. 1989. The continuum of plant responses to herbivory: the influence of plant associations, nutrient availability, and timing. Am. Nat., 134:1-19.
MARQUIS, D. A. 1974. The impact of deer browsing on Allegheny hardwood regeneration. U.S. Department of Agriculture. Forest Service Research Paper NE-308, Upper Darby, Pennsylvania, USA.
--. 1981. Effect of deer browsing on timber production in Allegheny hardwood forests of northwestern Pennsylvania. U.S. Department of Agriculture. Forest Service Research Paper NE475, Broomall, Pennsylvania, USA.
MYERS, J. A., M. VELLEND., S. GARDESCU AND P. L. MARKS. 2004. Seed dispersal by white-tailed deer: implications for long-distance dispersal, invasion, and migration of plants in eastern North America. Oecologia, 139:35-44.
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION (NOAA). 2004. National Climatic Data Center. Climatography of the United States. No. 20. 1971-2000. http://cdo.ncdc.noaa.gov/climatenormals/ clim20/il/119029.pdf. Accessed 12 Nov. 2007.
PASCHKE, J. E. AND J. D. ALEXANDER. 1970. Soil survey of Lake County Illinois. U.S. Soil Conservation Service, Urbana, Illinois, USA.
PRIMACK, R. B., S. L. MIAO AND K. R. BECKER. 1994. Costs of reproduction in the pink lady's slipper orchid (Cypripedium acaule): defoliation, increased fruit production, and fire.
ROONEY, T. P. AND D. M. WALLER. 2003. Direct and indirect effects of white-tailed deer in forest ecosystems. For. Ecol. Manage., 181:165-176.
ROYO, A. A., R. COLLINS, M. B. ADAMS, C. KIRSCHBAUM AND W. P. CARSON. 2010a. Pervasive interactions between ungulate browsers and disturbance regimes promote temperate forest herbaceous diversity. Ecol., 91:93-105.
--, S. L. STOUT, D. S. DECALESTA AND T. G. PIERSON. 2010b. Restoring forest herb communities though landscape-level; deer herd reductions: is recovery limited by legacy effects? Biol. Conserv., 143:2425-2434.
RUHREN, S. AND S. N. MANDEL. 2003. Herbivory constrains survival, reproductions and mutualisms when restoring nine temperate forest herbs. J. Torrey Bot. Soc., 130:34-42.
RUSSELL, F. L., D. B. ZIPPIN AND N. L. FOWLER. 2001. Effects of white-tailed deer (Odocoileus virginianus) on plants, plant populations and communities: a review. Am. Mid. Nat., 146:1-26.
RUZICKA, K.J.,J. w. GRONINGER AND J.J. ZAZCEK. 2010. Deer browsing, forest edge effects, and vegetation dynamics following bottomland forest regeneration. Rest. Ecol., 18:702-710.
SAS INSTITUTE. 2004. Version 9.1.3. SAS Institute, Cary, North Carolina, USA.
SHANNON, C. E. AND W. WEAVER. 1949. The mathematical theory of communication. The University of Illinois Press, Urbana, USA.
SIMPSON, E. H. 1949. Measurement of species diversity. Nature, 163:688.
SOKAL, R. R. AND F. J. ROHLF. 1995. Biometry: the principles and practice of statistics in biological research. W. H. Freeman, New York, New York, USA.
STROLE, T. A. AND R. C. ANDERSON. 1992. White-tailed deer browsing: species preferences and implications for central Illinois forests. Nat. Areas J., 12:139-144.
SUDING, K. N., K. L. GROSS AND G. R. HOUSEMAN. 2004. Alternative states and positive feedbacks in restoration ecology. Trends Ecol. Evol., 19:46-53.
SUZUKI, M., T. MIYASHITA, H. KABAYA, K. OCHIAI, M. ASADA AND T. TANGE. 2008. Deer density affects ground-layer vegetation differently in conifer plantations and hardwood forests on the Boso Peninsula, Japan. Ecol. Res., 23:151-158.
SWINK, F. AND G. WILHELM. 1994. Plants of the Chicago region. Fourth edition. Indiana Academy of Science, Indianapolis, Indiana, USA.
THOMAS, L., S. T. BUCKLAND, E. A. REXSTAD, J. L. LAAKE, S. STRINDBERY, S. L. HEDLEY, J. R. B. BISHOP, T. A. MARQUES AND K. P. BURNHAM. 2010. Distance software: design and analysis of distance sampling surveys for estimating population size. J App. Ecol., 47:5-14.
TREMBLAY, J., J. HUOT AND F. POTVIN. 2006. Divergent nonlinear responses of the boreal forest field layer along an experimental gradient of deer densities. Oecologia, 150:78-88.
TILGHMAN, N. G. 1989. Impacts of white-tailed deer on forest regeneration in northwestern Pennsylvania. J. Wild. Manage., 53:524-532.
UNITED STATES CENSUS BUREAU (USCB). 2009, http://quickfacts.census.gov/ qfd/states/17/17097.html. Accessed 16-08 2009.
URBANEK, R. E., C. K. NIELSEN, T. 8. PKEUSS AND G. A. GLOWACKI. 2011. Cost comparison of vegetation monitoring techniques used to assess white-tailed deer herbivory. Trans. IL. State Acad., 104:163-172.
--, --, -- AND --. 2012a. Comparison of aerial surveys and pellet-based distance sampling methods for estimating deer density. Wild. Soc. Bull., DOI:10.1002/wsb116
--, --, -- AND --. 2012b. White-tailed deer (Odocoileus virginanus) herbivory in herbaceous plant communities in northeastern Illinois. Nat. Areas J., 32:6-14.
VELLEND, M., J. A. MYERS, S. GARDESCU AND P. L. MARKS. 2003. Dispersal of Trillium seeds by deer: implications for long-distance migration of forest herbs. Ecol., 84:1067-1072.
WALLER, D. M., S. JOHNSON, R. COLLINS ANO E. WILLIAMS. 2009. Threats posed by ungulate herbivory to forest structure and plant diversity in the Upper Great Lakes Region--with a review of methods to assess those threats. U.S. Department of Interior, Natural Resource Report NPS/GLKN/ NRR-2009/102.
WEBB, W. L. 1959. Summer browse preferences of Adirondack white-tailed deer. J. Wild. Manage., 23:455-46.
WEBSTER, C. R. AND G. R. PARKER. 2000. Evalulation of Osmorhiza claytonia (Michx.) C. B. Clarke, Arisaema triphyllum (L.) Schott, and Actacea pachypoda Ell. as potential indicators of white-tailed deer overabundance. Nat. Areas J., 20:176-188.
--, M. A. JENKINS AND J. H. ROCK. 2005. Long-term response of spring flora to chronic herbivory and deer exclusion in Great Smoky Mountains National Park, USA. Biol. Conserv., 125:297-307.
WHIGHAM, D. F. 1990. The effect of experimental defoliation on the growth and reproduction of a woodland orchid, Tipularia discolor. Can. J. Bot., 68:1812-1816.
SUBMITTED 26 AUGUST 2010
ACCEPTED 26 OCTOBER 2011
(1) Corresponding author: e-mail: firstname.lastname@example.org; Telephone: (915) 480-6376
RACHAEL E. URBANEK (1)
Cooperative Wildlife Research Laboratory, Southern Illinois University, Car bondale 62901
CLAYTON K. NIELSEN
Cooperative Wildlife Research Laboratory, Southern Illinois University, Carbondale 62901 and
Department of Forestry, Southern Illinois University, Carbondale 62901
GARY A. GLOWACKI AND TIMOTHY S. PREUSS
Lake County Forest Preserve District, Libertyville, Illinois 60048
|Gale Copyright:||Copyright 2012 Gale, Cengage Learning. All rights reserved.|