Printer Friendly

Patterns of selective herbivory on five prairie legume species.

INTRODUCTION

North American tallgrass prairie has been reduced to less than 10% of its original range, mainly by conversion to cropland, and prairie ecosystems now primarily exist as a highly fragmented patchwork of roadside easements, protected areas, and restorations (Gibson, 2009). The native community of mammalian herbivores has been similarly disrupted, with the loss of bison (Bison bison) from most prairie restorations and native remnants (Sanderson et al., 2008), resulting in an increased functional role of small mammalian herbivores such as voles (Howe et al., 2006) and increased incursion by edge-loving or forest-dwelling herbivores such as white-tailed deer (Odocoileus virginianus; Oehler et al., 1995) and Eastern cottontail rabbits (Sylvilagus floridanus). It is well established that mammalian herbivores can have large impacts on plant communities (Ritchie and Tilman, 1995; Augustine and Frelich, 1998; Augustine and McNaughton, 1998; Ritchie et al., 1998; Knops et al., 2000; Cote et al., 2004) and the selectivity and extent of grazing by mammalian herbivores in prairie ecosystems have potential consequences for plant community composition (Howe et al., 2002).

Mammalian herbivores often graze selectively (Huntly, 1991; Pastor et al., 1993; Augustine & Frelich, 1998; Ritchie et al., 1998; Howe and Lane, 2004), with consequences for both plant community composition and soil nutrient availability (Milton, 1979; Huntly, 1991; Pastor et al., 1993; Ritchie et al., 1998; Kempel et al., 2011). Selective grazing may be based on several factors influencing plant digestibility or palatability, including plant protein, fiber, nutrient contents, and anti-herbivore defenses (Milton, 1979; Marquis and Batzli, 1989; Huntly, 1991; Kempel et al., 2011; Pastor et al., 1993; Ritchie et al., 1998). In particular, grazing pressure is positively associated with plant nitrogen (N) content (Ritchie and Tilman, 1995; Ritchie et al., 1998; Knops et al., 2000), and some herbivores may preferentially graze high-N tissue within an individual plant species and within individual plants (Sanford and Huntly, 2009).

Legumes, with their high N content and high forage quality compared to other prairie plants, are commonly selected by mammalian herbivores (Gibson, 2009). White-tailed deer (Odocoileus virginianus), Eastern cottontail rabbits (Sylvilagus floridanus), and meadow voles (Microtus pennsylvanicus) have all been shown to preferentially browse or clip legumes in prairie or savanna ecosystems (Dusi, 1952; Bailey, 1969; Ritchie and Tilman, 1995; Howe and Lane, 2004; Sullivan and Howe, 2009). Legume populations are often limited by herbivory and when plants are protected from herbivores by exclosures, some fast growing, early maturing legume species are able to dramatically increase in abundance, while the relative abundances of less preferred legume species (e.g., Lespedeza capilata and Amorpha canescens) do not change (Ritchie and Tilman, 1995; Ritchie et al., 1998; Knops et al., 2000). In exclosure studies deer and voles have been shown to significantly reduce abundances of preferred legume species (e.g., Lathyrus venosus Muhl. ex Willd. for deer and Desmodium canaden.se for voles) that would otherwise become dominant (Ritchie et al., 1998; Howe and Lane, 2004).

Fire also drives community dynamics in prairies by impacting the physical and chemical environment. Fire and herbivory have interactive effects on plant community structure and composition, through both positive and negative feedbacks (Collins, 1987; Fuhlendorf et al., 2009). Removal of biomass by grazing or selective foraging on inflammable species can decrease fire frequency, as high grass biomass is positively correlated with fire frequency (Hobbs, 2006). In turn, fire influences patterns of herbivore selection of forage sites, as plant nutrient content, digestibility, and the availability of forage plants can temporarily increase following burning (Carlson et al., 1993; Mixon et al., 2009). White-tailed deer have shown preference for recently burned forage sites (Main and Richardson, 2002); however, deer preference for burned areas is more strongly documented in forests than in prairies, and the response of deer and other mammalian herbivores to burning in prairies is not well known. Distance from forest edge can also impact herbivory from this suite of herbivores, as meadow vole herbivory has been shown to vary with distance from forest edge (Nickel et al., 2003), and white-tailed deer are considered edge species (Alverson et al, 1988; Oehler et al., 1995; Cote et al., 2004) that may utilize prairie sites close to forest edge more heavily than more distant sites.

In this study we quantified the extent of selective grazing by mammalian herbivores on five legume species common in both tallgrass prairie restorations and nearby prairie remnants; Amorpha canescens Pursh, Dalea Candida Michx. ex Willd., Dalea purpurea Vent., Desmodium canadense (L.) DC., and Lespedeza capitala Michx.. We also determined the patterns of selective herbivory in relation to burn history and distance from forest edges and analyzed plant tissue chemistry to attempt to explain patterns of herbivore preference. We hypothesized selectivity for certain species may differ even among legume species and patterns of herbivory would be strongly influenced by plant nitrogen content, tissue carbon (C) quality, recent burn history, and proximity to the prairie-forest boundary.

METHODS

STUDY SITE

Carleton College Cowling Arboretum in Northfield, Minnesota (arboretum office at 44[degrees]28'00.2"N, 93[degrees]08'56.9"W) consists of 360 ha adjacent to Carleton College and is comprised mainly of upland and floodplain forest and restored prairie. According to the General Land Survey conducted in the mid-19th century, the area historically consisted of a variety of different habitats including prairie, closed-canopy forest, oak savanna, and thickets at the prairie-forest border. With European settlement the area was converted into agricultural land for the cultivation of corn (Zea mays L.) and soybeans (Glycine max (L.) Merr.). Beginning in 1995 the Arboretum initiated a prairie restoration project, annually converting a portion of agricultural land to restored prairie every year from 1995-2008 (Camill et al., 2004; Hernandez et al., 2013). Each restoration block was broadcast seeded with seeds harvested from several neighboring native remnant prairies, primarily McKnight prairie, located 11 km east of the study site. The restored prairies are highly diverse, with approximately 35 species with >1% cover in most planting blocks. The current restoration consists of a continuous prairie made up of 14 planting blocks that are similar in land use and restoration history, seed bank, and abiotic factors but vary in time since restoration (Fig. 1; for a more complete description of soil characteristics in these blocks see Hernandez et al., 2013). The northern and western edges of the prairie are bordered by forest and oak savanna, the eastern edge (east of the 2007 and 2008 planting blocks) is bordered by agricultural land, and Minnesota Highway 19 borders the southern edge.

Blocks used in this study were restored in 1996, 1998, 2001, 2004, 2006, and 2008. Blocks were chosen based on their burn history, proximity to the prairie-forest boundary, and abundance of legume species. Arboretum prairies are managed by burning, which occurs on a 4-year rotational cycle. Blocks 2004, 2008, and the southern halves of blocks 1996 and 2001 were burned in the spring of 2012, the spring of the growing season during which transect sampling occurred (Fig. 1). Remaining blocks were considered unburned for this analysis, as they had not been burned during this growing season. The 1998 and 2006 blocks were burned in the spring of 2009 and the northern halves of the 1996 and 2001 blocks were burned in the spring of 2011 (Nancy Braker, pers. comm.).

Deer and other mammalian herbivores are common in the region. Although deer were nearly extirpated from the late 1800s to the 1920s, their population has recovered and increased steadily since then (Augustine and Frelich, 1998). Average prefawn deer density in 2011 in the state management unit that includes the Arboretum was 3.7 deer/[km.sup.2] (Minnesota DNR). Deer populations are managed in the Arboretum by an annual winter archery hunt. Although meadow voles are the most common small mammals observed in trapping surveys, their abundance is not known. Rabbit abundance is also unknown.

FIELD SAMPLING

Six 25 X 2 m transects were established in each of the 1996, 1998, 2001, 2004, and 2006 planting blocks of the Carleton College Arboretum between 11 and 31 Jul. 2012. Eight transects were established in the 2008 block to account for low legume densities. Transects were spaced evenly throughout each block and the direction of each transect (north-south or east-west) depended on the dimensions of the block. Due to differences in block size, not all transects were equally distant from one another, but the distance between transects was always greater than 25 m. Herbivore damage and plant characteristics were recorded for the five legume species. The total numbers of individuals sampled of each species across all blocks were: Amorpha (171), D. Candida (527), D. purpurea (465), Desmodium (298), and Lespedeza (936). Average density (number of individuals per square meter) for each legume species are shown for each block in Table 1. Variables measured included plant height, total number of stems, and the number of clipped stems, indicating the number of stems that were damaged by mammalian herbivores.

In Jul. 2013 (the growing season following the transect sampling) we harvested individuals of each plant species to determine how tissue C and N content and C quality varied among species and with respect to grazing and recent burning (i.e., burned in Spring 2013). We targeted 20 individuals of each species--10 from recently burned areas (five grazed and five ungrazed) and 10 from unburned areas (five grazed and five ungrazed). On ungrazed plants we harvested the top 20 cm of all stems from each individual plant. On grazed plants we harvested the top 20 cm of only those stems that had been ungrazed as a best estimate of the tissue that would commonly be consumed by herbivores. We did not find any grazed Amorpha, therefore n = 10 for this species in our analyses (five burned, five unburned). Similarly, Lespedeza is commonly found with a single stem, therefore it was difficult to find grazed plants with ungrazed stems. For Lespedeza n = 16 (five burned/ungrazed, five unburned/ungrazed, four burned/grazed, two unburned/ungrazed).

PLANT TISSUE CHEMISTRY

Harvested plant tissue was dried at 65 C and ground in a Wiley Mill (Thomas Scientific, Swedesboro, New Jersey). We used a subsample of ground plant tissue to measure three C fractions--soluble cell contents, hemicellulose and bound proteins, and recalcitrant contents (cellulose, lignin, and other recalcitrants)--using an Ankom 200 fiber analyzer (Ankom Technology, Fairport, New York). Samples were washed with solutions (neutral detergent solution, acid detergent solution, and a 72% sulfuric acid solution) to dissolve particular C fractions and dried and weighed between washes to determine the proportion of each fraction. Samples were damaged in the final extraction, therefore it was not possible to distinguish the cellulose fraction from the other recalcitrant materials. To measure plant C and N concentrations, we combined all remaining plant tissue in each treatment and finely ground samples with a mixer mill (CertiPrep 8000-D; SPEX SantplePrep, Metuchen, New Jersey). One composite sample was analyzed for each species in each treatment--burned/grazed, burned/ungrazed, unburned/ungrazed, and unburned/grazed (with the exception of Amorpha, for which grazed individuals were not collected). Samples were analyzed for total C and N by combustion on an ECS 4010 element analyzer (Costech Analytical, Valencia, California).

STATISTICAL ANALYSIS

An electivity index was used to determine herbivore preference for each species by ascertaining whether herbivory occurred at a rate predicted by species abundance (random selection; Jenkins, 1979; De Jager et al., 2013):

[E.sub.i] = ln [[r.sub.i](1 - [p.sub.i])/[p.sub.i](1 - [r.sub.i])]

In this equation, [r.sub.i] is the proportion of grazed stems from species i, and [p.sub.i] is the proportion of all stems from species i relative to the total number of stems from all surveyed plants. Positive [E.sub.i] values indicate preference while negative [E.sub.i] values indicate avoidance. [E.sub.i] values were calculated for all individuals of each species, and [E.sub.i] values were also calculated separately for burned and unburned individuals of each species to see if there was a difference in herbivore selectivity in response to burning. Statistical significance for each [E.sub.i] value was determined using a corresponding chi-square test with one degree of freedom (Jenkins, 1979; De Jager et al., 2013):

[X.sup.2] = [E.sub.i.sup.2]/(1/[x.sub.i] + 1/m - [x.sub.i]) + (1/[y.sub.i] + 1/n - [y.sub.i])

where [x.sub.i] is the number of grazed stems from species i, [y.sub.i] is the total number of stems from species i, m is the number of all grazed stems across the five species surveyed and n is the total number of stems surveyed.

For each plant surveyed, percent grazed was calculated by dividing the number of stems that exhibited herbivore damage by the total number of stems. t-tests were used to characterize differences in average percent grazed between bunted and unburned fields for each species surveyed. Density of each legume species was calculated at the transect level and correlations were used to determine the relationship between legume density and grazing.

The distance of each transect to the nearest forest edge was calculated by using a GIS map incorporating GPS coordinates of the approximate starting location of each transect and a GIS layer with polygons representing the forest edge. Transect orientation depended on block size and shape, and transects were sometimes parallel and sometimes perpendicular to forest edge. Transects ranged from 25-480 m from the nearest forest edge. Distance was calculated by using the near tool in ArcGIS. For each species Pearson correlation tests were used to determine the relationship between distance and corresponding transect-level average percent grazed. Because the two species of Dalea are closely related and frequently co-occur across the restoration, we combined them for this analysis.

We determined differences in plant tissue chemistry using an MANOVA that included tissue C, N, and C fractions. Comparisons among species for each constituent of tissue chemistry were performed using a three-way ANOVA with plant species, burning, and grazing as our fixed factors. Each C fraction was analyzed separately. Plant C:N was compared among species using a one-way ANOVA with plant species as the fixed factor. In all cases, post-hoc mean comparisons were made using Tukey's HSD.

RESULTS

Herbivore selectivity was not uniform across legume species, and for all species grazing was significantly different from what would be expected to occur at random, indicating that species are either significantly selected or avoided (Fig. 2; Table 2). Desmodium was the most highly preferred species with the highest percent grazed, (39% [+ or -] 2.3, [E.sub.i] = 1.50, P < 0.001) followed by D. Candida (26% [+ or -] 1.7, [E.sub.i] = 0.49, P < 0.001) and D. purpurea (31% [+ or -] 1.9, [E.sub.i] = 0.21, P < 0.01). Lespedeza and Amorpha were both avoided, with [E.sub.i] values of -1.65 (P < 0.001) and -1.75 (P < 0.001) and average percentage grazed of 6% [+ or -] 0.06 and 2% [+ or -] 1.0, respectively. The average percent of plants that exhibited any level of herbivore damage followed the same pattern as average percent of stems grazed, with 2.9% for Amorpha, 34.3% for D. Candida, 45.2% for D. purpurea, 58.1% for Desmodium, and 10.6% for Lespedeza.

Average percent grazed was higher in burned plots compared to unburned plots for three species: D. Candida (burned = 45.78% [+ or -] 8.28; unburned = 24.90% [+ or -] 1.74; P < 0.01), D. purpurea (burned = 51.52% [+ or -] 3.06; unburned = 15.95% [+ or -] 1.86; P < 0.001), and Lespedeza (burned = 13.23% [+ or -] 2.38; unburned = 4.80% [+ or -] 0.64; P < 0.001; Fig. 3). There was no difference in grazing between burned and unburned plots for Desmodium or Amorpha (P = 0.584 and 0.387, respectively). We calculated [E.sub.i] values between burned and unburned plots (Table 2). D. Candida was grazed at expected rates based on its abundance in unburned plots ([E.sub.i] = 0.004) but was significantly preferred in burned plots ([E.sub.i] = 2.963, P < 0.001). D. purpurea was avoided in unburned plots ([E.sub.i] = -0.607, P < 0.001) and was preferred in burned plots ([E.sub.i] = 0.623, P < 0.001). Lespedeza was grazed at rates expected by its abundance in burned plots ([E.sub.i] = -0.249, P > 0.15) but was avoided in unburned plots ([E.sub.i] = -2.155, P < 0.001). In contrast with other species, Desmodium-was more highly preferred in unburned plots ([E.sub.i] = 1.548, P < 0.001) than burned plots ([E.sub.i] = 0.480, P < 0.001). Avoidance of Amorpha was strong across all plots but was slightly lower in burned plots ([E.sub.i] = -1.341, P < 0.001) compared to unburned plots ([E.sub.i] = -1.740, P < 0.001).

There was a marginally significant negative correlation between average percent grazed and distance from the woods for the two Dalea species (r = -0.298, P = 0.092). However, there was no significant relationship between distance from the woods and average percent grazed for Amorpha (P = 0.427), Desmodium (P = 0.608), or Lespedeza (P = 0.258). For all five legume species, rate of grazing on a particular species was not affected by the density of that species (Amorpha, P = 0.606; both Dalea species, P = 0.265; Desmodium, P = 0.503; Lespedeza, P = 0.174).

There were highly significant differences in legume tissue chemistry among species when considering all aspects of tissue chemistry measured (MANOVA: Pillai = 3.094, [F.sub.24,44] = 6.259, P < 0.001). Plant C quality was significantly different among plant species (three-way ANOVA; P < 0.001; Table 3). There was a significant effect of plant species on all C fractions, but the main effects of burning or grazing on any C fraction was not significant (three-way ANOVA; P > 0.05). There was a significant two-way interaction between burning and grazing for the hemicellulose and protein fraction (P = 0.017), with this fraction being greater for ungrazed plants collected from burned blocks and greater for grazed plants collected in unburned blocks (data not shown). There was a significant two-way interaction between species and grazing for the recalcitrant fraction (P = 0.004) and a significant three-way interaction (species X burning X grazing) for the soluble C fraction (P = 0.014) and hemicellulose and protein fraction (P = 0.043). These significant interactions were generally due to differences in only one or two of the five species and the species driving the response was different for each C fraction.

Considering only the main effect of species, Amorpha had a significantly greater concentration of the recalcitrant C fraction compared to other species, while D. Candida and D. purpurea had the two highest concentration of the hemicellulose and protein fraction and the two lowest concentrations of the recalcitrant fraction (Tukey's HSD; P < 0.05). Desmodium and Lespedeza had the two highest concentrations of cell solubles (Tukey's HSD; P < 0.05). Plant C:N ratios also varied among species (Table 3). Lespedeza had the lowest %N and the highest C:N of all of the species (one-way ANOVA; Tukey's HSD).

DISCUSSION

HERBIVORE SELECTIVITY AMONG LEGUME SPECIES

Herbivores exhibited high selectivity among legume species, as both [E.sub.i] and average percent grazed varied greatly: highly preferred Desmodium exhibited an average of 39% of stems grazed while on average only 2% of Amorpha exhibited herbivore damage. In addition recent burning increased the rate of grazing on some species while having no effect on others, and this effect was independent of the degree of selectivity among legume species. Several studies have shown that deer and other mammalian herbivores exhibit high selectivity for legumes over other plants (Huntly, 1991; Knops et al., 2000; Spotswood et at., 2002; Cote et al., 2004). However, our results indicate there is high selectivity even within legumes and fire may differentially affect the rate of grazing on legume species.

We suspect deer are the primary grazers of Desmodium and Lespedeza, while the two Dalea species are susceptible to herbivory from all mammalian herbivores. In a related study in the Carleton Arboretum using exclosures that exclude deer while admitting small mammalian herbivores including rabbits and voles, both Dalea species were commonly grazed inside the deer exclosures while Desmodium was not (D. Hernandez, pers. obs.). In addition the location of grazing on Desmodium and Lespedeza was frequently greater than 30 cm from the ground, above the reach of small mammals, while Dalea were commonly grazed within 10 cm of the soil (A. Nisi, pers. obs.). Desmodium and Lespedeza are relatively woody and thick-stemmed when mature, while Dalea are much smaller with thinner stems. For some plants the impact of vole and rabbit herbivory is restricted to early spring growth and decreases throughout the growing season due to increased stem toughness and wideness (Bailey, 1969; Howe, 2008), and we suspect this is true for both Desmodium and Lespedeza. Amorpha, a woody-stemmed shrub-like species, was rarely eaten, with only 3% of all plants exhibiting any herbivore damaged and an average of 2% of total stems grazed. This high avoidance is consistent with other studies (Ritchie et al., 1998) and since grazing events were so infrequent we are unable to hypothesize which herbivores are responsible for Amorpha grazing.

Legumes are highly selected by herbivores because N-fixation by rhizobial symbionts gives legumes relatively high tissue N compared to other plant functional groups (Tjoelker et al., 2005; Novotny et al., 2007). However, N content alone is not sufficient to explain the high degree of selectivity observed between legume species, and previous studies have shown similar C:N ratios among the five legume species in our study (Tjoekler et al., 2005; Novotny et al., 2007). Our results suggest herbivore preference and avoidance are related to both tissue N content and C quality. Amorpha had a C:N ratio and %N similar to that of Desmodium but had a significantly higher proportion of recalcitrant C, which likely offset the relatively high N content in terms of herbivore digestibility. Conversely, the recalcitrant C fraction of Lespedeza was similar to Desmodium, but Lespedeza had significantly lower N content than other species. Therefore, while Amorpha was likely avoided based on recalcitrant C content, Lespedeza was likely avoided because of low tissue N content relative to other prairie legumes.

EFFECTS OF DISTANCE FROM FOREST EDGE

Contrary to our hypotheses, there was not a strong effect of distance to the prairie-forest boundary on the rate of grazing. It is possible over a larger scale, herbivory levels would decrease as distance from the forest increases, hut that our study area was not large enough to see these effects. The maximum distance of our transects from the forest edge (480 m) may he within the range where edge effects still occur. Alternatively, rabbits and deer may not be restricted to the forest. Particularly in the case of rabbits, prairie may provide them as much protection as forest, and one study (Bond et al., 2002) found rabbits spend a substantial amount of time both in forest and grasslands. It is likely herbivory by rabbits and deer is common in restored prairies, as most prairie patches now occur in isolated fragments, rather than the large expanses of grassland that once characterized native tallgrass prairie. Additionally, some herbivores species may avoid areas near the forest edge. For example meadow vole herbivory has been shown to be greater in locations away from the prairie-forest edge (Nickel et al., 2003). In the present study we were unable to distinguish herbivory among mammal species, which would have allowed us to determine how grazing patterns of each herbivore are uniquely affected by the proximity to forest. Nevertheless, proximity to the forest edge does not seem to have a significant effect on grazing of legume species in our study site.

BURNING EFFECTS ON SELECTIVITY

Recent burning increased the rate of grazing on Dalea and Lespedeza but had no effect on the other legume species. This effect was independent of the distance from the forest edge, and there was no relationship between plant density and average percent grazed for any plant, indicating that the effects of density, distance, and burning on average percent grazed for these legume species are independent of one another. In contrast to other species, Desmodium was more highly preferred in unburned than burned areas (Table 2). However, the electivity index is a measure of relative preference and the average percent grazed of Desmodium did not differ between burned and unburned areas. Rather, the lower preference for Desmodium in burned areas was a function of higher herbivore preference for Dalea relative to Desmodium.

Herbivores have been shown to preferentially graze recently burned areas (Biondini et al., 1999; Main & Richardson, 2002). Fires in early spring remove detritus and increase the amount of solar radiation to the soil, resulting in earlier emergence of shoots, which are then susceptible to herbivory (Knapp and Seastedt, 1986; Meek et al., 2008). Fire also increases the availability of non-N nutrients, like phosphorus, that may be limiting resources to legume growth (Knapp and Seastedt, 1986; Ritchie and Tilman, 1995; Roscher et al., 2011). Therefore, fire may make legumes more palatable, more visible to herbivores early in the season, or both.

Since there was no significant difference in plant C quality or N content within a species in burned and unburned blocks, increased preference for burned plants is likely due to factors other than increased palatability following fire. It is likely that increased visibility following the removal of biomass and earlier emergence of shoots were responsible for the increased grazing pressure. However, it is unclear why grazing did not increase on all legume species. One possibility is that burning disproportionately affects the visibility of shorter plants that are otherwise not visible to mammalian herbivores, as may be the case for Dalea. Alternatively, shifts in plant emergence time in response to burning may differ among species, although differences in emergence time of our target species are unknown. Nevertheless, recent burning may have significant species-specific effects on herbivore preference, although it appears to affect only those species that are moderately preferred or avoided, as grazing on the most preferred species, Desmodium, and the most avoided, Amorpha, was not affected by burning.

IMPLICATIONS FOR PRAIRIE RESTORATION

Selectivity among legume species and species-specific effects of prescribed burning have important implications for prairie restoration and management. Herbivory on legume species may alter plant community composition, and consequently N mineralization and cycling, by reducing or eliminating species that contribute nitrogen to the ecosystem through fixation (Knops et al., 2000). Although the relative rates of N fixation among legume species in our study are unknown, the selective grazing of species with higher tissue N concentrations or lower concentrations of recalcitrant C have the potential to impact rates of C and N cycling in this system. In addition prescribed burning, a common management tool in prairie restoration, is used in part to increase the richness and evenness of desirable species such as legumes (Towne and Knapp, 1996; Bowles et al., 2003; Bowles and Johnson, 2013). However, through interactions with herbivory, fire may have differential effects on plant species, even within the same functional group. Furthermore, selective herbivory on legumes may interact with burning to negatively impact soil N concentrations (Knops et al., 2010). Taken together, these findings suggest the interactive effects of fire and herbivory must be considered to determine the ultimate impacts of management on plant community composition.

Acknowledgments.--Special thanks to Teddy Gelderman and the Northfield Middle School TORCH students that assisted with data collection in the field. We thank Nancy Braker and Matt Elbert for logistical support and their work on the restoration and management of the Arboretum prairies, and Greg Phillips for GIS help. Collection of plant tissue and analysis of carbon fractions was performed by Jared Beck, Hannah Tremblay, and Eleanor Youngblood. Nathan De Jager provided helpful comments on the manuscript. Funding was provided by the National Science Foundation (DEB# 1021194).

SUBMITTED 2 AUGUST 2013

ACCEPTED 4 AUGUST 2014

LITERATURE CITED

ALVERSON, W. S., D. M. WALLER, AND S. L. SOLHEIM. 1988. Forests too deer: edge effects in northern Wisconsin. Conserv. Biol, 2:348-358.

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 S. J. MCNAUGHTON. 1998. Ungulate effects on the functional species composition of plant communities: herbivore selectivity and plant tolerance. J. Wildlife Manage., 62:1165-1183.

BAILEY, J. A. 1969. Exploratory study of nutrition of young cottontails. J. Wildlife Manage, 33:346-&.

BOND, B. T., L. W. BURGER, AND M. D. JONES. 2013. Repeated burning of eastern tallgrass prairie increases richness and diversity, stabilizing late successional vegetation. Ecol. Appl., 23:464-478.

--, --, B. D. LEOPOLD, J. C. JONES, AND K. D. GODWIN. 2002. Habitat use by cottontail rabbits across multiple spatial scales in Mississippi./. Wildlife Manage., 66:1171-1178.

BOWLES, M. L., M. D. JONES, AND J. L. MCBRIDE. 2003. Twenty-year changes in burned and unburned sand prairie remnants in northwestern Illinois and implications for management. Am. Midi. Nat., 149:35-45.

CAMILL, P., M. MCKONE, S. T. STURGES, W. J. SEVERUD, E. ELLIS, J. LIMMER, C. B. MARLIN, R. T. NAVRATIL, A. J. PURDIE, B. S. SANDEL, S. TALUKDER, AND A. TROUT. 2004. Community- and ecosystem-level changes in a species-rich tallgrass prairie restoration. Ecol. Appl., 14:1640-1694.

CARLSON, P. C., G. W. TANNER, J. M. WOOD, AND S. R. HUMPHREY. 1993. Fire in key deer habitat improves browse, prevents succession, and preserves endemic herbs. J. Wildlife Manage., 57:914-928.

COLLINS, S. L. 1987. Interaction of disturbances in tallgrass prairie--a field experiment. Ecology, 68:1243-1250.

COTE, S. D., T. P. ROONEY, J. P. TREMBLAY, C. DUSSAULT, AND D. M. WALLER. 2004. Ecological impacts of deer overabundance. Annu. Rev. Ecol. Evol. Syst., 35:113-147.

DE JAGER, N. R., B. J. COGGER, AND M. A. THOMSEN. 2013. Interactive Effects of Flooding and deer (Odocoileus virginianus) browsing on floodplain forest recruitment. Forest Ecol. Manag., 303:11-19.

DUSI, J. L. 1952. The Food Habits of several populations of cottontail rabbits in Ohio. J. Wildlife Manage., 16:180-186.

FUHLENDORF, S. D., D. M. ENGLE, J. KERBY, AND R. HAMILTON. 2009. Pyric herbivory: rewilding landscapes through the recoupling of fire and grazing. Conserv. Biol., 23:588-598.

GIBSON, D. J. 2009. Grasses and grassland ecology. Oxford University Press, New York.

HERNANDEZ, D. L., E. H. ESCH, C.J. ALSTER, M. J. MCKONE, AND P. CAMILL. 2013. Rapid acculumlation of soil carbon and nitrogen in a prairie restoration chronosequence. Soil Sci. Am. J., 77:2029-2038.

HOBBS, N. T. 2006. Large herbivores as sources of disturbance in ecosystems, p. 261-288. In: K. Danell, P. Duncan, L. Bergstrom, and J. Pastor (eds.). Large herbivore ecology, ecosystem dynamics and conservation. Campbridge University Press, Cambridge, U.K.

HOWE, H. F. 2008. Reversal of fortune: plant suppression and recovery after vole herbivory. Oecologia, 157:279-286.

--, J. S. BROWN, AND B. ZORN-ARNOLD. 2002. A rodent plague on prairie diversity. Ecol. Lett., 5:30-36.

-- AND D. LANE. 2004. Vole-driven succession in experimental wet-prairie restorations. Ecol. Appl., 14:1295-1305.

--, B. ZORN-ARNOLD, A. SULLIVAN, AND J. S. BROWN. 2006. Massive and distinctive effects of meadow voles on grassland vegetation. Ecology, 87:3007-3013.

HUNTLY, N. 1991. Herbivores and the dynamics of communities and ecosystems. Annu. Rev. Ecol. Syst., 22:477-503.

JENKINS, S. H. 1979. Seasonal and year-to-year differences in food selection by beavers. Oecologia, 44:112-116.

KEMPEL, A., M. SCHADLER, T. CHROBOCK, M. FISCHER, AND M. VAN KLEUNEN. 2011. Tradeoffs associated with constitutive and induced plant resistance against herbivory. Proc. Natl. Acad. Sci., 108: 5685-5689.

KNAPP, A. K. AND T. R. SEASTEDT. 1986. Detritus accumulation limits productivity of tallgrass prairie. Bioscience, 36:662-668.

KNOPS, J. M. H., M. E. RITCHIE, AND D. TILMAN. 2000. Selective herbivory on a nitrogen fixing legume (Lathyrus venosus) influences productivity and ecosystem nitrogen pools in an oak savanna. Ecoscience, 7:166-174.

MARQUIS, R. J. AND G. O. BATZLI. 1989. Influence of chemical factors on palatability of forage to voles. J. Mammal., 70:503-511.

MAIN, M. B. AND L. W. RICHARDSON. 2002. Response of wildlife to prescribed fire in southwest Florida pine flatwoods. Wildlife Soc. B., 30:213-221.

MEEK, M. G., S. M. COOPER, M. K. OWENS, R. M. COOPER, AND A. L. WAPPEL. 2008. White-tailed deer distribution in response to patch burning on rangeland. J. Arid Environ., 72:2026-2033.

MILTON, K. 1979. Factors influencing leaf choice by howler monkeys--test of some hypotheses of food selection by generalist herbivores. Am. Nat., 114:362-378.

MIXON, M. R., S. DEMARAIS, P. D. JONES, AND B. J. RUDE. 2009. Deer forage response to herbicide and fire in mid-rotation pine plantations. J. Wildlife Manage., 73:663-668.

NICKEL, A. M., B. J. DANIELSON, AND K. A. MOLONEY. 2003. Wooded habitat edges as refugia from microtine herbivory in tallgrass prairies. Oikos, 100:525-533.

NOVOTNY, A. M., J. D. SCHADE, S. E. HOBBIE, A. D. KAY, M. KYLE, P. B. REICH, AND J. J. ELSER. 2007. Stoichiometric response of nitrogen-fixing and non-fixing dicots to manipulations of CO2, nitrogen, and diversity. Oecologia, 151:687-696.

OEHLER, M. W., J. A. JENKS, AND R. T. BOWYER. 1995. Antler rubs by white-tailed deer--the importance of trees in a prairie environment. Can. J. Zool., 73:1383-1386.

PASTOR, J., B. DEWEY, R. J. NAIMAN, P. F. MCINNES, AND Y. COHEN. 1993. Moose browsing and soil fertility in the boreal forests of Isle Royale National Park. Ecology, 74:467-480.

RITCHIE, M. E. AND D. TILMAN. 1995. Responses of legumes to herbivores and nutrients during succession on a nitrogen-poor soil. Ecology, 76:2648-2655.

--, --, AND J. M. H. KNOPS. 1998. Herbivore effects on plant and nitrogen dynamics in oak savanna. Ecology, 79:165-177.

ROSCHER, C., S. THEIN, A. WEIGELT, V. M. TEMPERTON, N. BUCHMANN, AND E. D. SCHULZE. 2011. N-2 fixation and performance of 12 legume species in a 6-year grassland biodiversity experiment. Plant Soil, 341:333-348.

SANDERSON, E. W., K. H. REDFORDA, B. WEBER, K. AUNE, D. BALDES, J. BERGER, D. CARTER, C. CURTIN, J. DERR, S. DOBROTT, E. FEARN, C. FLEENER, S. FORREST, C. GERLACH, C. C. GATES, J. E. GROSS, P. GOGAN, S. GRASSEL, J. A. HILTY, M. JENSEN, K. KUNKEL, D. LAMMERS, R. LIST, K. MINKOWSKI, T. OLSON, C. PAGUE, P. ROBERTSON, AND B. STEPHENSON. 2008. The ecological future of the North American Bison: conceiving long-term, large-scale conservation of wildlife. Conserv. Biol., 22:252-266.

SANFORD, M. P. AND N. HUNTLY. 2009. Selective herbivory by the desert woodrat (Neotoma Lepida) on Joshua trees (Yucca Brevifolia). West N. Am. Nat., 69:165-170.

SPOTSWOOD, E., K. L. BRADLEY, AND J. M. KNOPS. 2002. Effects of herbivory on the reproductive effort of 4 prairie perennials. BMC Ecol., 2:2.

SULLIVAN, A. T. AND H. F. HOWE. 2009. Prairie forb response to timing of vole herbivory. Ecology, 90:1346-1355.

TJOELKER, M. G., J. M. CRAINE, D. WEDIN, P. B. REICH, AND D. TILMAN. 2005. Linking leaf and root trait syndromes among 39 grassland and savannah species. New Phytol., 167:493-508.

TOWNE, E. G. AND A. K. KNAPP. 1996. Biomass and density responses in tallgrass prairie legumes to annual fire and topographic position. Am. J. Bot., 83:175-179.

ANNA C. NISI, DANIEL L. HERNANDEZ, (1) LYDIA P. ENGLISH and EMILY S. ROGERS

Department of Biology, Carleton College, 1 North College Street, Northfield, Minnesota 55057

(1) Corresponding author: e-mail: dhernand@carleton.edu

TABLE 1.-Average plant density (plants [m.sup.-2]) of legume
species in each block. Average densities were calculated by
averaging transect-level densities (number of plants in 50
[m.sup.2] transect) and se denotes standard error of average
densities

            A. canescens       D. Candida

Block   Density     SE     Density     SE

1996    0.4667    0.1640   0.0000    0.0000
1998    0.0733    0.0434   0.0033    0.0033
2001    0.0033    0.0033   0.4300    0.4300
2004    0.0000    0.0000   0.0233    0.0233
2006    0.0100    0.0045   1.2367    0.4421
2008    0.0125    0.0084   0.0475    0.0262

            D. purpurea       D. canadense

Block   Density     SE     Density     SE

1996    0.5900    0.3738   0.5933    0.1739
1998    0.4700    0.2917   0.2100    0.0781
2001    0.1767    0.1767   0.0300    0.0113
2004    0.0000    0.0000   0.0833    0.0233
2006    0.2500    0.1039   0.0733    0.0272
2008    0.0475    0.0251   0.0025    0.0025

           L. capitata

Block   Density     SE

1996    0.3167    0.1359
1998    2.3200    0.5998
2001    0.3433    0.1411
2004    0.0067    0.0042
2006    0.0333    0.0152
2008    0.0750    0.0403

TABLE 2.--Electivity indices of the five legume species with
respect to recent burning. Asterisks indicate those cases
where the rate of grazing was significantly different (P <
0.001) from what its abundance would suggest (i.e., [E.sub.i]
significantly different from 0)

           A. canescens    D. Candida    D. purpurea

Burned     -1.341 *        2.963 *       0.623 *
Unburned   -1.740 *        0.004         -0.607 *

           D. canadense    L. capitata

Burned     0.480 *         -0.249
Unburned   1.548 *         -2.155 *

TABLE 3.--Plant tissue carbon and nitrogen concentrations
and carbon fractions for the Five legume species (means with
SE in parentheses). Letters denote significant differences
among species within a column

Species                %C              %N

A. canescens   48.47 (1.25) a   2.98 (0.36) a
D. Candida     43.00 (0.05) b   3.17 (0.07) a
D. purpurea    42.91 (0.11) b   3.08 (0.08) a
D. canadense   45.18 (0.15) c   2.92 (0.10) a
L. capitata    47.10 (0.27) a   2.31 (0.07) b

Species               C:N             Cell solubles

A. canescens   16.45 (1.57) a    50.45 (0.83)   a
D. Candida     13.57 (0.27) b    45.27 (1.18)   b
D. purpurea    13.94 (0.32) ab   50.85 (1.42)   a
D. canadense   15.53 (0.55) ab   59.14 (0.39)   c
L. capitata    20.47 (0.61) c    55.49 (1.07)   c

               Hemicellulose    Cellulose, lignin,
Species        and proteins     and recalcitrants

A. canescens   15.13 (0.46) a   35.06 (0.54) a
D. Candida     33.72 (1.15) b   21.62 (0.26) b
D. purpurea    24.21 (1.49) c   25.57 (0.38) c
D. canadense   12.26 (0.39) a   29.20 (0.31) d
L. capitata    14.12 (0.27) a   30.99 (0.98) d
COPYRIGHT 2015 University of Notre Dame, Department of Biological Sciences
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2015 Gale, Cengage Learning. All rights reserved.

 
Article Details
Printer friendly Cite/link Email Feedback
Author:Nisi, Anna C.; Hernandez, Daniel L.; English, Lydia P.; Rogers, Emily S.
Publication:The American Midland Naturalist
Article Type:Report
Geographic Code:1USA
Date:Jan 1, 2015
Words:6394
Previous Article:Tebuthiuron use in restoring degraded tallgrass prairies and warm-season grass pastures.
Next Article:Effects of grassland biomass harvest on nesting pheasants and ducks.
Topics:

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters