SEARCH

SEARCH BY CITATION

Keywords:

  • AIC ;
  • Ecological fidelity;
  • FQI ;
  • Mean C;
  • Nachusa Grasslands;
  • Net primary productivity;
  • Non-metric multidimensional scaling;
  • Soil

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Questions

Was ecological fidelity (structure/composition, function and durability) restored in a series of tallgrass prairie restorations? Which factors influenced success? Can success be assessed in prairie restoration using indices of ecological fidelity?

Location

Nachusa Grasslands in Lee and Ogle Counties, Illinois, USA.

Methods

To assess restoration of ecological fidelity, mean coefficient of conservatism (Mean C), floristic quality index (FQI), above-ground net primary productivity (ANPP), soil bulk density, total soil nitrogen (N) and total soil carbon of 19 restoration sites were recorded from 20 × 50 m modified Whittaker plots across a chronosequence, and compared to benchmark values acquired from both the literature and field observation of remnant prairies. Following assessment, multiple factors were examined through correlation analysis, Akaike's information criterion and multiple regression analysis to determine the relationship of these factors to restoration success.

Results

All restoration sites attained the benchmark value for Mean C, while only four attained the benchmark value for FQI. Mean C and FQI both decreased across the chronosequence. Frequency of prescribed fire and soil bulk density had significant positive relationships to Mean C. FQI was best explained by the FQI value of the seed mix sown. Thirteen restoration sites attained the benchmark value for ANPP, which remained stable across the restoration chronosequence. Abundance of exotic species and soil drainage had a negative relationship to ANPP. Few restoration sites attained benchmark values for soil bulk density, total N and total carbon, and none of the sites showed a trajectory towards benchmark values across the chronosequence.

Conclusions

Our study demonstrates that high-quality seed mixes may aid in establishing prairie restorations with high scores of floristic quality. However, restoration of vegetation does not guarantee the successful restoration of ecological function. Long-term monitoring is needed to more effectively assess durability and the multiple factors that influence restoration quality. Overall, the three components of ecological fidelity related to structure/composition, function and durability provide a useful framework to assess restoration success and guide management. Our study can serve as a model for future research and assessment of restoration success.


Nomenclature
Swink & Wilhelm (1994)

 

Abbreviations
ANPP

annual net primary productivity

FQA

floristic quality assessment

FQI

floristic quality index

Mean C

mean coefficient of conservatism

N

nitrogen

NMDS

non-metric multidimensional scaling

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

On a global scale, ecological restoration has become a popular practice to recover lost ecosystem attributes in order to meet conservation objectives and to return the flow of ecological goods and services to human populations (Clewell & Aronson 2007). In central North America, the prairie ecosystem has been reduced in extent more than any other ecosystem on the continent, from ca. 162 to 29 million ha, a decrease of more than 80% (Samson & Knopf 1994). In addition, conversion of remaining prairie for human use has greatly outpaced protection (Hoekstra et al. 2005). Therefore, ecological restoration will be critical for the future persistence of this once extensive ecosystem and its associated species, processes and services.

Ecological restoration of prairie has been exemplified by a number of high profile projects, e.g. the Curtis Prairie in Madison, Wisconsin (Kucharik et al. 2006), as well as numerous grassroots-organized projects throughout the Great Plains region (Packard & Mutel 2005). However, many restoration projects are initiated without a clear vision of desired outcomes (Herrick et al. 2006; Fagan et al. 2008) and few established projects are thoroughly evaluated to determine if success is being achieved (Ruiz-Jaen & Aide 2005). Furthermore, defining and evaluating restoration success are two concepts that are rarely articulated clearly (Hobbs & Harris 2001; Zedler 2007).

One potential way to define and evaluate restoration success is to use a method based on the concept of ‘ecological fidelity’ (Higgs 1997). For the three criteria of ecological fidelity to be met a restoration must: (1) exhibit a similar species structure/composition relative to the original ecosystem; (2) function similarly to the original ecosystem; and (3) be durable (i.e. criteria 1 and 2 must persist over time). Ecological function and durability are often overlooked in restoration evaluations (Ruiz-Jaen & Aide 2005); thus, most evaluations of prairie restoration have not been ecologically comprehensive.

Evaluations based solely on species structure and/or composition have found restoration success to be limited. Species richness of restored prairies is usually less than that of remnant prairies, and richness often declines in a restored prairie over time (Sluis 2002; Camill et al. 2004; McLachlan & Knispel 2005). Conversely, exotic species richness is generally higher in restored prairie than in remnant sites (Martin et al. 2005; McLachlan & Knispel 2005).

Studies that have included other measures beyond species structure and/or composition also have shown limited restoration success. For example, soil properties of restored prairie such as bulk density, total nitrogen (N), and total carbon slowly recover over time, but not consistently to levels of remnant prairie, even after many years (Baer et al. 2002; Brye et al. 2002; McLachlan & Knispel 2005). However, levels of annual net primary productivity (ANPP) similar to or higher than remnant prairie have been frequently observed (Baer et al. 2002; Camill et al. 2004; Martin et al. 2005).

The objectives of this study were to determine the: (1) success of a large-scale prairie restoration project at Nachusa Grasslands in northern Illinois, USA, by measuring the ecological fidelity of 19 restoration sites and comparing measurements to specific benchmarks of restoration success; (2) relationship between restoration success and factors such as management history, seed-mix quality and site characteristics; and (3) sets of ecological attributes that would be most suitable to assess ecological fidelity. We focused on a comprehensive set of attributes often examined in ecosystem function and restoration.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Site description

Nachusa Grasslands, owned and managed by The Nature Conservancy (TNC), consists of ca. 1275 ha of land ca. 135 km west of Chicago in north-central Illinois, US (41°89′ N, 89°34′ W) (Fig. 1). Mean annual temperature is 8.5 °C, mean annual precipitation is 94.7 cm (61.2 cm rain from April through September), and during the primary year of this study (2005) the mean temperature was 11.5 °C and the area received 75.3 cm of precipitation (20.8 cm during the growing season; ISWS 2012).

image

Figure 1. Sample site locations of 19 restored prairies and two row-crop fields in The Nature Conservancy's Nachusa Grasslands in northern Illinois, USA, indicating where modified Whittaker plots were placed. Numerical site labels refer to management unit identification numbers, SF refers to Soybean Field and CF refers to Corn Field (see Table S1-1 in Appendix S1).

Download figure to PowerPoint

Nachusa Grasslands is comprised of mixed prairies on scattered hills with a thin underlying layer of glacial till (see Appendix S1 for additional site details). The mixture of vegetation includes a mosaic of 11 natural community types, including mesic tallgrass prairie, gravel hill prairie, oak savanna and various wetland types (Jones & Cushman 2004). Dominant species in the tallgrass areas include the grasses Andropogon gerardii, Sorgastrum nutans and Andropogon scoparius, and the forbs Solidago rigida, Ratibida pinnata and Lespedeza capitata. At Nachusa Grasslands, some of the remnant hill prairies and wetlands remain in a high-quality biological state because they were unsuitable for cultivation. However, virtually all the mesic prairie in the areas between hill prairie and wetland was pastured by 1873 and converted to row-crop production by 1939. These areas are now the focus of the restoration efforts at Nachusa.

The Nature Conservancy made their first land acquisition at Nachusa in 1986. At that time, the land was primarily comprised of hill prairie remnants and row-crop fields in the lower-lying areas. Restoration of fields to prairie began in 1987. At the time of our sampling, restoration had been completed on 61 sites, ranging in size from <1 ha to >14 ha. The restoration sites were intended to be species-rich tracts of native vegetation connecting the remnants within the preserve to form a large, contiguous ecosystem. Methods of restoration have varied over time and have included hand and mechanical sowing of seed collected from the study site and local remnants, as well as seed purchased from local nurseries. Additional management has included prescribed fire, supplemental seed addition and exotic species removal. In this study, 19 restoration sites were evaluated. The sites were of similar soil type, restored 2–19 yrs ago and from 1.13–14.88 ha in size.

Species structure/composition

From 30 June to 14 August 2005, vegetation was surveyed within a 20 × 50 m modified Whittaker plot (Barnett & Stohlgren 2003) randomly placed in each of the 19 restoration sites. To minimize edge effects, plots were not placed within 20 m of a site's edge. Species presence/absence was recorded for the entire 1000-m2 plot. Within each of the ten 1-m2 subplots, the abundance of each species was measured by visually estimating percentage canopy based on the modified Daubenmire scale (Abrams & Hulbert 1987).

Presence/absence and cover data were used to calculate species richness, Simpson's index of diversity (1-D, where D = Σpi2, and pi2 is the relative cover of individuals of the ith species). The presence/absence data were also used in a floristic quality assessment (FQA; Swink & Wilhelm 1994) using the mean coefficient of conservatism (Mean C) and floristic quality index (FQI) (Appendix S2). Only native species were included in our calculations of Mean C and FQI. The FQI of the seed mix for each restoration site was also calculated. FQA is a useful method for measuring prairie vegetation quality (Taft et al. 2006; McNicoll & Augspurger 2010; Sivicek & Taft 2011; Spyreas et al. 2012), despite the subjective nature of assigning C values to species and a potential bias towards rare species (Bowles & Jones 2006; Landi & Chiarucci 2010).

Function – above-ground net primary productivity

When tall, warm-season perennial grasses were flowering in 2005 (late August to early September), vegetation was clipped from a 0.1-m2 quadrat randomly placed within each of the ten 1-m2 subplots in each restoration site. All vegetation was clipped to a 2-cm level. Live vegetation was separated from previous years’ dead vegetation (i.e. detritus), forbs were separated from grasses, and native from non-native species (distinguished according to Swink & Wilhelm 1994). Clipped material was oven-dried at 60 °C for 48 h. For each restoration site, dried live material from the ten 0.1-m2 quadrats was combined and weighed to determine above-ground net primary productivity (ANPP) as g·m−2 for each restoration site.

Function – soil properties

Between July and September 2005, a cylindrical soil core (2-cm diameter and 15-cm depth) was collected from each of the ten 1-m2 subplots in each restoration site. Each core was split into two portions consisting of 0–5 and 5–15 cm depths. For each restoration site, ten cores were combined into a single composite sample for each of the two depths. Ten soil cores were also collected from one corn (Zea mays) field and one soybean (Glycine max) field in which modified Whittaker plots were randomly placed. The two fields were located at Nachusa and represent ‘restoration sites’ at time zero in the chronosequence. Additionally, ten soil cores were collected from modified Whittaker plots randomly placed in three small mesic prairie remnants. Two of the remnants were located at Nachusa and the third was ca. 15 km away along a railroad right-of-way (Lee County: 41°76′ N, 89°39′ W). The area of modified Whittaker plots was decreased by 50% due to the small size of the remnants.

Soil cores were kept at ca. 4 °C in an ice chest and transported to Southern Illinois University-Carbondale. The soil cores were passed through a 2-mm sieve and refrigerated at ca. 4 °C until analysis. Soil pH and electrical conductivity were measured using a ratio of 2 g soil to 15 ml ultrapure water (Blake 1965) with a Fischer Scientific Accumet Basic AB15 pH Meter (Thermo Fischer Scientific, Inc., Waltham, MA, US) and a Mettler-Toledo MC126 Conductivity Meter (Mettler-Toledo, Inc., Columbus, OH, US). Two samples of ca. 50 μg of oven-dried (50 °C for 1 wk) finely milled soil were analysed for total soil N and carbon (Sollins et al. 1999) for each depth and restoration site using a Thermo FlashEA 1112 N and carbon analyser (Elantech, Lakewood, NJ, US). For each soil variable, the means of both depths for each plot were used in data analysis.

During April 2006, three bulk density cores were collected from each plot (a single core from each of the plots’ two 10-m2 and single 100-m2 subplots) using a 4.8-cm diameter soil corer and hammer attachment. Each core was split into two portions consisting of 0–5 and 5–15 cm depths. Cores were weighed before and after drying at 105 °C for 1 wk. The mean of the three bulk density values was calculated for each site and used to compare total N and carbon values of restoration sites, row-crop fields and prairie remnants based on an equivalent mass of soil.

Benchmarks of restoration success

A nearby mesic remnant of high vegetative quality was not available for comparison to the restored sites for the floristic quality and ANPP analyses. Although Nachusa Grasslands has several high-quality remnant hill prairies, their floristic characteristics were different from the mesic, lowland sites where restoration has taken place, thus making comparison inappropriate. Therefore, to gauge restoration success of floristic quality and ANPP, comparisons utilized benchmark values from the literature instead of on-site reference prairies. The three remnant prairies described earlier were, however, used as reference sites for the comparison of soil properties.

Floristic quality values of individual restoration sites were compared to benchmark values based on Swink & Wilhelm's (1994) method of identifying natural area quality in plant communities (Appendix S2). Accordingly, a plant community scoring a Mean C ≥ 3.5 or an FQI ≥ 35 was considered to be of at least marginal natural area quality. Based on those benchmarks identified by Swink & Wilhelm (1994), if a restoration site scored a Mean C ≥ 3.5 or an FQI ≥ 35, the restoration site was deemed successful in terms of plant species structure/composition. Mean C and FQI indices perform differently in certain circumstances (Taft et al. 1997, 2006), therefore we used both indices in order to make a more comprehensive assessment.

The ANPP values were compared to a benchmark value based on the summary of Risser et al. (1981) ANPP values from throughout the North American tallgrass prairies. Their values provided a benchmark range (mean ± 1 SD) of 296–499 g·m−2 (= 23, mean = 397.5 g·m−2); if a restoration site's ANPP fell within that range, the restoration site was considered successful.

Values of soil properties acquired from the three remnant prairies were used to develop a benchmark range based on their means ± 1 SD. If the value of a restoration site's soil property fell within the benchmark range of remnant prairies, the restoration site was considered successful.

Durability

To determine the durability of the restoration sites, values of floristic quality, ANPP and soil properties for each restoration site were compared to restoration site age using linear regression (PROC GLM) in SAS (v. 9.1.3; SAS Institute Inc., Cary, NC, US) Restoration sites were considered durable if values showed a trajectory towards, or maintenance of, the benchmark value.

Factors influencing restoration success

A three-step statistical analysis was followed to examine relationships between restoration success in terms of Mean C, FQI, ANPP and total soil N and a priori sets of independent variables (Appendix S3). The three steps were (1) correlation analysis, (2) Akaike's information criterion (AIC) model selection, and (3) multiple regression analysis (see Appendix S3 for details) and were conducted in SAS.

Slope and soil drainage values were acquired from Ogle (Acker et al. 1980) and Lee (Zwicker 1985) County Soil Surveys. Weather data were acquired from the State Climatologist Office for Illinois (ISWS 2012). Distances from the centre of each modified Whittaker plot to the nearest prairie remnant, restoration site, road and service lane were calculated using a global positioning system and the spatial analysis tools in ArcMap (v. 9.1; Environmental Systems Research Institute (ESRI), Redlands, CA, US).

Non-metric multidimensional scaling

Non-metric multidimensional scaling (NMDS) ordination was performed using DECODA (Minchin 1990) to compare the species presence/absence observed in each restoration site to the seed mix sown at each site. The ordination used one to four dimensions, 100 random starting configurations, 200 maximum iterations, a stopping value for stress reduction of 0.9999, a stress stopping value of 0.01 and the Bray–Curtis coefficient dissimilarity measure. The number of dimensions retained for interpretation was determined through examination of scree plots and stress values (McCune & Grace 2002).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Species structure/composition

All 19 restoration sites evaluated had Mean C scores ≥ 3.5, indicating that the restoration of plant species composition was successful (Fig. 2a). However, only four of the 19 restoration sites evaluated had FQI scores ≥ 35, indicating that the restoration of plant species composition was not always successful with respect to that metric (Fig. 2b). Species richness and Simpson's index of diversity were both highly correlated with FQI (r = 0.86, < 0.0001 and r = 0.66, = 0.0019, respectively; Appendix S4), therefore only FQI results are reported here.

image

Figure 2. Chronosequence analyses for restoration sites at The Nature Conservancy's Nachusa Grasslands. Values of mean coefficient of conservatism (Mean C) (symbols for two sites at 7 yrs after restoration have similar values and thus overlap) (a), floristic quality index (FQI) (b), and annual net primary productivity (ANPP) (c); and chronosequence analyses including comparison to two agricultural fields (at age zero) and three remnant prairies for soil bulk density (d), total soil N (e), and total soil C (f). Broken lines represent benchmark level or range, indicative of a successful restoration. Linear regressions of the variables vs restoration age were significant for Mean C (F1,17 = 8.60, = 0.0093, R2 = 0.34) and FQI (F1,17 = 11.70, = 0.0033, R2 = 0.41). Linear regressions of soil bulk density, total soil N and total soil carbon were not significant (> 0.05).

Download figure to PowerPoint

Function

Thirteen of the 19 restoration sites had ANPP within the benchmark range for tallgrass prairie (Fig. 2c). For soil properties, only results from the 0–5 cm depth are reported here because they were highly correlated with the 5–15 cm depth (total soil N: r = 0.94, < 0.0001; total soil carbon: r = 0.93, < 0.0001; soil bulk density: r = 0.76, < 0 .0001). One restoration site exhibited bulk density levels within the range of the remnant prairies, while bulk density of the other restoration sites was well above the upper limit of the range (Fig. 2d). For total soil N and carbon, only two (Fig. 2e) and one (Fig. 2f) restoration site, respectively, exhibited levels within the benchmark range of the remnant prairies. All other restoration sites were below the lower limits of the respective ranges.

Durability

Among the attributes evaluated, only ANPP exhibited durability according to our chronosequence analyses (Fig. 2c). Chronosequence analyses suggested that Mean C (Fig. 2a) and FQI (Fig. 2b) declined sharply with restoration site age, while soil bulk density (Fig. 2d), total soil N (Fig. 2e) and total soil carbon (Fig. 2f) remained stable but below benchmark levels, suggesting durability had not been achieved in the restoration sites.

Factors influencing restoration success

For Mean C, four variables – fire frequency, soil bulk density, soil pH and distance to nearest remnant prairie (Appendix S3) – were retained for multiple regression analysis. Of these, fire frequency and soil bulk density had significant relationships (both positive) to Mean C in the model (Table 1a).

Table 1. Multiple regression results of factors influencing restoration success vs independent variables in the Nachusa Grasslands: (a) mean coefficient of conservatism (Mean C), (b) floristic quality index (FQI), (c) annual net primary productivity (ANPP), and (d) total soil N (complete list in Appendix S3) including parameter estimates (PE) and standardized estimates (SE)
Independent VariablePESE P
(a) Mean C (F4,11 = 6.75, = 0.0054, R2 = 0.71)
Fire Frequency1.060.600.0045
Soil Bulk Density1.840.560.0075
Soil pH−0.53−0.240.2643
Distance to Nearest Remnant Prairie0.00−0.070.7593
(b) FQI (F5,10 = 5.58, < 0.0104, R2 = 0.74)
Seed Mix FQI0.460.670.0276
Sample Date0.120.240.2251
Fire Frequency4.860.170.5049
Distance to Nearest Remnant Prairie0.000.090.6760
Distance to Nearest Road/Service Lane0.000.000.9991
(c) ANPP (F7,11 = 4.7, = 0.0115, R2 = 0.75)
Exotic Biomass%−421.26−0.790.0021
Soil Drainage−55.27−0.590.0381
Total Soil N1.840.370.0831
Slope14.280.470.0894
Soil Bulk Density−316.72−0.350.1308
Soil pH86.680.24ּ0.1485
Planting FQI−1.09−0.070.7256
(d) Total Soil N (F4,14 = 4.98, = 0.0104, R2 = 0.59)
Native C4 Grass% Cover−2.73−2.710.0251
Native Forb% Cover−252.87−2.500.0355
Planting FQI−1.71−0.490.0729
Exotic C3 Grass% Cover−180.61−1.620.0995

For FQI, five variables – seed mix FQI, sample date, distance to nearest remnant prairie, distance to nearest road and fire frequency (Appendix S3) – were retained for multiple regression analysis. Of these, only seed mix FQI had a significant relationship (positive) to FQI of the restoration sites in the model (Table 1b). Seed mix FQI was subsequently examined across a chronosequence to determine if younger restoration sites were receiving seed mixes of higher quality relative to older ones, but no pattern was detected related to age of restoration (F1,14 = 4.35, = 0.0557, R2 = 0.24).

For ANNP, seven variables were retained for multiple regression analysis – percentage of exotic ANPP, total soil N, soil drainage, soil bulk density, slope, soil pH and FQI – of the restoration sites (Appendix S3). Of these, the percentage of exotic ANPP and soil drainage had significant relationships (both negative) to ANPP of the restoration sites in the model (Table 1c).

For total soil N, four variables – native C4 grass cover, native forb cover, FQI of the restoration site and exotic C3 grass cover (Appendix S3) – were retained for multiple regression analysis. Of these, native C4 grass cover and native forb cover had significant relationships (both negative) to total soil N in the model (Table 1d).

Non-metric multidimensional scaling

Two dimensions (stress = 0.164) were retained for the ordination solution comparing species presence/absence observed in each restoration site to species presence/absence in the original seed mix of a restoration site (Fig. 3). Seven restoration sites with high FQI values of ca. 35 formed a well-defined group in the centre of the ordination. The ordination plot implies that restoration sites established using seed mixes comprised of similar species were not always similar in terms of observed species composition; sometimes, similar restoration sites were established from mixes of dissimilar species composition. Seed mixes were aligned on the left side of the ordination sample plot and restoration sites were aligned on the right side (Fig. 3), indicating differences in composition of the seed mixes and composition of the restoration sites.

image

Figure 3. Non-metric multidimensional scaling ordination of species composition of the seed mix connected to the observed species composition of the restoration sites at The Nature Conservancy's Nachusa Grasslands. Circle encloses the seven sites with the highest observed FQI values (four sites with FQI > 35 (successful restoration) and three sites with FQI ca. 35).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Comprehensive assessment of grassland restoration success is necessary to determine if restoration targets are being met and to allow restoration ecologists and land managers the opportunity to develop methods of best practice. Our study assessed ecological fidelity to determine the restoration success of numerous prairie restorations. Using this approach, we showed that some indicators are more sensitive than others to the assessment of restoration success. Comparable studies of restoration have shown the value of assessing multiple indicators of ecological fidelity, including attributes related to vegetation (Matthews et al. 2009), spatial structure (Seabloom & Van Der Valk 2003) and soil microbes (Smith et al. 2003). Nevertheless, few examine multiple indicators in the same way as undertaken in our study. The chronosequence approach taken here is an effective method for measuring ecosystem changes over time (Foster & Tilman 2000; Knops & Tilman 2000; Baer et al. 2002), although factors other than age can result in observed differences among sites (Pickett 1989; Johnson & Miyanishi 2008; Walker et al. 2010).

Restoration composition and durability

Many studies have found that ecological fidelity often decreases with time after restoration. The declines in Mean C and FQI observed in the chronosequence of this study were consistent with studies of prairie restorations (Sluis 2002; Camill et al. 2004; McLachlan & Knispel 2005) and remnants (Bowles & Jones 2006). Nevertheless, our study suggests higher values of Mean C and especially FQI among the younger restoration sites relative to the older restoration sites may indicate that prairie restoration characterized by high floristic quality is possible in the early stages of restoration. This observation is important, given the previous implication that conservative species contributing to high FQI values may take decades to establish (Schramm 1992), and prairie restorations rarely achieve FQI levels of natural areas (Taft et al. 1997).

A decrease in metric values used to describe species composition of prairie restorations is often attributed to the tendency of C4 grasses to become dominant at the expense of subdominant native forbs (Sluis 2002; Camill et al. 2004; McLachlan & Knispel 2005), which comprise the majority of species with high C-values. Although C4 grass cover increased with restoration site age in our study, native forb cover remained constant. However, both richness and Mean C of native forbs decreased significantly with restoration site age even though cover remained constant, suggesting that C4 grass expansion may have negatively affected native forb composition.

Seed limitation may account for low diversity in prairie restoration (Martin & Wilsey 2006). A high level of species richness of the seed mix used in a prairie restoration can increase the number of species that eventually become established (Polley et al. 2005; Piper et al. 2007). Our findings corroborate these studies, as restoration site FQI success was linked to seed mixes of high FQI value (and thus species richness). Increasing seed mix richness for a restoration site increases the available species pool for that restoration, and larger species pools contribute to greater species richness (Taylor et al. 1990; Partel & Zobel 1999). Increasing seed mix richness may aid certain species to overcome dispersal limitations, thus increasing the likelihood that these species will become established (Tilman 1997).

While seed mix quality seemed to have a strong influence on Mean C and FQI values in prairie restorations, plant composition and floristic quality were likely affected by additional factors. Anecdotal evidence suggests that those restoration sites at Nachusa that received a more species-rich seed mix could have also received additional management to improve and maintain floristic quality. For example, over-seeding (Foster et al. 2007; Williams et al. 2007), mowing (Williams et al. 2007) and increased seed density at time of planting (Sheley & Half 2006; Piper et al. 2007) can all increase species establishment in prairie restorations. However, although implemented at Nachusa, none of these common management activities were tested for an effect on restoration success in this study because accurate records of these activities were not available.

In our study, higher frequency of prescribed fire and high soil bulk density displayed strong positive relationships with high Mean C values. These results were somewhat unexpected, as frequent fire favours C4 grasses (Kucera & Koelling 1964; Hulbert 1986; Gibson & Hulbert 1987) and high bulk density is associated with poorer drainage and thus a wetter prairie. Among remnants, evidence suggests no relationship between fire frequency and Mean C values (although a positive relationship with FQI was reported), and drier prairies have higher Mean C values than mesic prairies (Bowles & Jones 2006). Correlation analysis indicated a significant positive relationship between soil bulk density of the restoration and seed mix quality (Appendix S3, Table S3-2) so that seed mix quality may have been positively influencing Mean C values, as it did with FQI values, although soil bulk density performed better in the model.

Restoration function and durability

Durability is important to maintain stability of grassland restorations. Our results were consistent with other studies that have shown ANPP levels can be restored to durable levels characteristic of tallgrass prairie (Baer et al. 2002; Camill et al. 2004; Martin et al. 2005). However, although durability was being achieved, younger restoration sites exhibited successful production levels more consistently than older restoration sites, suggesting that ANPP destabilizes as restoration sites age, thus threatening durability. Older restoration sites also exhibited lower Mean C and FQI values than younger restoration sites. ANPP also had a negative relationship with exotic species abundance, primarily Poa compressa, P. pratensis and Bromus inermis. When these species were abundant, native C4 grasses that comprise the majority of ANPP in prairies were excluded, resulting in decreased ANPP. Early spring growth and dominance of Poa and Bromus may suppress seedling establishment of native C4 grasses (Foster 1999), negatively effecting ANPP (Smith & Knapp 1999). Soil drainage also had a negative relationship with ANPP, which was expected, as decreased soil moisture is often associated with decreased production in the eastern tallgrass prairie (Sala et al. 1988; Briggs & Knapp 1995). However, other factors such as fire regime, total soil N and light availability can all interact with moisture in dictating production levels from year-to-year (Knapp et al. 1998), but were unrelated to ANPP in our study, perhaps because the prevalence of exotic grasses overrode their influence.

As in other studies, soil characteristics in the restoration sites were different from the remnants, likely reflecting their cultivation history (Deluca & Keeney 1993; Knops & Tilman 2000; Baer et al. 2000; Baer et al. 2002). For example, soil bulk densities in the prairie restoration sites of this study were ca. 25% higher than in the remnants, and total soil N and carbon levels were ca. 75% and 50% of the remnants, respectively. Variation in total soil N and carbon levels among the remnants was also higher relative to the restored sites. Cultivation can lead to increased soil bulk density due to compaction from heavy farm equipment (Miller & Gardiner 1998) and soil consolidation from weathering (Potter et al. 1999). We did not detect a recovery in these soil characteristics across the chronosequence, although levels of total soil N and carbon can accrue slowly and can be difficult to detect in restored prairies (Burke et al. 1995; Knops & Tilman 2000; Camill et al. 2004; McLachlan & Knispel 2005). A restored prairie may take ca. 150 yrs before levels of soil carbon reach those of remnant prairie (Potter et al. 1999); the oldest restoration site we sampled was <20 growing seasons old. Additionally, few N-fixing legumes were observed in the restoration sites, and no relationship between legume abundance and total soil N was found.

Management implications

From a management perspective, ecological fidelity is important as a comprehensive measure of the success in reaching restoration targets. The success of restoring ecological fidelity at Nachusa Grasslands was mixed, depending on which component of fidelity was considered. When considering vegetation composition, floristic quality in terms of Mean C was restored in all restoration sites, but in only four sites in terms of FQI. When considering ecological function, ANPP was restored quite frequently, but soil properties rarely achieved benchmark levels. Success was limited for durability. Of the six measures evaluated in this study, only ANPP was considered durable. However, the initial success of some of the younger restorations at Nachusa is noteworthy, particularly those with high FQI values. Those restoration sites provide an indication that the complex species composition of prairie can be restored, at least in the short term, and therefore we encourage the use of seed mixes rich in species of high C-value if high floristic quality is a goal of restoration. We stress that all restoration methods and management activities, especially seeding densities at the time of restoration, should be thoroughly quantified at the time of their implementation so they can be subsequently evaluated and replicated if deemed effective.

This study underscores the importance of including ecologically comprehensive evaluations of restoration success in management plans for any type of ecosystem. Evaluations that focus strictly on vegetation composition are valuable (e.g. Matthews et al. 2009) but ecologically incomplete. To aid both current and future ecological restoration efforts, contributions from research and management should include comprehensive evaluation of the components of ecological fidelity. Ecosystems are inherently dynamic, so that to capture the temporal variability of composition, structure, function and gain additional insight on how that variability affects the durability of restoration projects, long-term monitoring will be essential to effective assessment (Bakker et al. 1996; Meyer et al. 2010). Each ecological attribute examined in this study was suitable as a measure of an ecological fidelity component, but other indicators would also be appropriate and insightful (e.g. insects, mammals, birds, available soil N, below-ground biomass). The approach of our evaluation can serve as a model for future research and assessment of restoration success.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

We thank Sara Baer, Loretta Battaglia, Rose Bloise, Cody Considine, Joe Hansen, Chris Hauser, Nate Hill, Bob Hoyle, Ryan Klopf, Peter Minchin, Tom Mitchell, Austin Saylor, Katie Schoenfeldt, Laura Shirley, Greg Spyreas, Jay Stacey and Mary Vieregg for assistance; Nachusa Grasslands Project Director Bill Kleiman for providing his expertise and access to the site; and The Nature Conservancy for financial support.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Abrams, M.D. & Hulbert, L.C. 1987. Effect of topographic position and fire on species composition in tallgrass prairie in northeast Kansas. American Midland Naturalist 117: 442445.
  • Acker, L.L., Hodges, M.S., Keller, G.T. & Rehner, R. 1980. Soil Survey of Ogle County, Illinois. National Cooperative Soil Survey. United States Department of Agriculture, Washington, DC, US.
  • Baer, S.G., Kitchen, D.J., Blair, J.M. & Rice, C.W. 2002. Changes in ecosystem structure and function along a chronosequence of restored grasslands. Ecological Applications 12: 16881701.
  • Baer, S.G., Rice, C.W. & Blair, J.M. 2000. Assessment of soil quality in fields with short and long term enrollment in the CRP. Journal of Soil and Water Conservation 55: 142146.
  • Bakker, J.P., Olff, H., Willems, J.H. & Zobel, M. 1996. Why do we need permanent plots in the study of long-term vegetation dynamics? Journal of Vegetation Science 7: 147156.
  • Barnett, D.T. & Stohlgren, T.J. 2003. A nested-intensity design for surveying plant diversity. Biodiversity and Conservation 12: 255278.
  • Blake, C. 1965. Methods of Soil Analysis, Part 2. American Society of Agronomy, Madison, WI, US.
  • Bowles, M. & Jones, M. 2006. Testing the efficacy of species richness and floristic quality assessment of quality, temporal change, and fire effects in tallgrass prairie natural areas. Natural Areas Journal 26: 1730.
  • Briggs, J.M. & Knapp, A.K. 1995. Interannual variability in primary production in tallgrass prairie: climate, soil moisture, topographic position, and fire as determinants of aboveground biomass. American Journal of Botany 82: 10241030.
  • Brye, K.R., Norman, J.T. & Gower, S.T. 2002. Assessing the progress of a tallgrass prairie restoration in southern Wisconsin. American Midland Naturalist 148: 218235.
  • Burke, I.C., Lauenroth, W.K. & Coffin, D.P. 1995. Soil organic matter recovery in semiarid grasslands: implications for the Conservation Reserve Program. Ecological Applications 5: 793801.
  • Camill, P., McKone, M.J., Sturges, S.T., Severud, W.J., Ellis, E., Limmer, J., Martin, C.B., Navratil, R.T., Purdie, A.J., Sandel, B.S., Talukder, S. & Trout, A. 2004. Community- and ecosystem-level changes in a species-rich tallgrass prairie restoration. Ecological Applications 14: 16801694.
  • Clewell, A.F. & Aronson, J. 2007. Ecological Restoration: Principles, Values, and Structure of an Emerging Profession. Island Press, Washington, DC, US.
  • Deluca, T.H. & Keeney, D.R. 1993. Soluble organics and extractable nitrogen in paired prairie and cultivated soils of central Iowa. Soil Science 155: 219228.
  • Fagan, K.C., Pywell, R.F., Bullock, J.M. & Marrs, R.H. 2008. Do restored calcareous grasslands on former arable fields resemble ancient targets? The effect of time, methods and environment on outcomes. Journal of Applied Ecology 45: 12931303.
  • Foster, B.L. 1999. Establishment, competition and the distribution of native grasses among Michigan old-fields. Journal of Ecology 87: 476489.
  • Foster, B.L. & Tilman, D. 2000. Dynamic and static views of succession: testing the descriptive power of chronosequence approach. Plant Ecology 146: 110.
  • Foster, B.L., Murphy, C.A., Keller, K.R., Aschenbach, T.A., Questad, E.J. & Kindscher, K. 2007. Restoration of prairie community structure and ecosystem function in an abandoned hayfield: a sowing experiment. Restoration Ecology 15: 652661.
  • Gibson, D.J. & Hulbert, L.C. 1987. Effects of fire, topography and year-to-year climatic variation on species composition in tallgrass prairie. Vegetatio 72: 175185.
  • Herrick, J.E., Schuman, G.E. & Rango, A. 2006. Monitoring ecological processes for restoration projects. Journal for Nature Conservation 14: 161171.
  • Higgs, E.S. 1997. What is good ecological restoration? Conservation Biology 11: 338348.
  • Hobbs, R.J. & Harris, J.A. 2001. Restoration Ecology: repairing the Earth's ecosystems in the new millennium. Restoration Ecology 9: 239246.
  • Hoekstra, J.M., Boucher, T.M., Ricketts, T.H. & Roberts, C. 2005. Confronting a biome crisis: global disparities of habitat loss and protection. Ecology Letters 8: 2329.
  • Hulbert, L.C. 1986. Fire effects on tallgrass prairie. In: Clambey, G.K. & Pemble, R.H. (eds.) Proceedings of the 9th North American Prairie Conference, pp. 138142. North Dakota State University, Fargo, ND, US.
  • ISWS (Illinois State Water Survey) 2012. [Illinois Climate]. www.isws.illinois.edu/atmos/statecli.
  • Johnson, E.A. & Miyanishi, K. 2008. Testing the assumptions of chronosequences in succession. Ecology Letters 11: 419431.
  • Jones, S.R. & Cushman, R.C. 2004. A Field Guide to the North American Prairie. Houghton Mifflin Co., New York, NY, US.
  • Knapp, A.K., Briggs, J.M., Blair, J.M. & Turner, C.L. 1998. Patterns and controls of aboveground net primary production in tallgrass prairie. In: Knapp, A.K., Briggs, J.M., Hartnett, D.C. & Collins, S.L. (eds.) Grassland Dynamics: Long-Term Ecological Research in Tallgrass Prairie, pp. 193221. Oxford University Press, New York, NY, US.
  • Knops, J.M.H. & Tilman, D. 2000. Dynamics of soil nitrogen and carbon accumulation for 61 years after agricultural abandonment. Ecology 81: 8898.
  • Kucera, C.L. & Koelling, M. 1964. The influence of fire on composition of central Missouri prairie. American Midland Naturalist 72: 142147.
  • Kucharik, C.J., Fayram, N.J. & Cahill, K.N. 2006. A paired study of prairie carbon stocks, fluxes, and phenology: comparing the world's oldest prairie restoration with an adjacent remnant. Global Change Biology 12: 122139.
  • Landi, S. & Chiarucci, A. 2010. Is floristic quality assessment reliable in human-managed ecosystems? Systematics and Biodiversity 8: 269280.
  • Martin, L.M. & Wilsey, B.J. 2006. Assessing grassland restoration success: relative roles of seed addition and native ungulate activities. Journal of Applied Ecology 43: 10981109.
  • Martin, L.M., Moloney, K.A. & Wilsey, B.J. 2005. An assessment of grassland restoration success using species diversity components. Journal of Applied Ecology 42: 327336.
  • Matthews, J.W., Spyreas, G. & Endress, A.G. 2009. Trajectories of vegetation-based indicators used to assess wetland restoration progress. Ecological Applications 19: 20932107.
  • McCune, B. & Grace, J.B. 2002. Analysis of Ecological Communities. MJH Software Design, Gleneden Beach, OR, US.
  • McLachlan, S.M. & Knispel, A.L. 2005. Assessment of long-term tallgrass prairie restoration in Manitoba, Canada. Biological Conservation 124: 7588.
  • McNicoll, M.B. & Augspurger, C.K. 2010. A comparison of vegetation and seed bank community structure in a sand prairie in Illinois, U.S.A. American Midland Naturalist 164: 136150.
  • Meyer, C.K., Whiles, M.W. & Baer, S.G. 2010. Plant community recovery following restoration in temporally variable riparian wetlands. Restoration Ecology 18: 5264.
  • Miller, R.W. & Gardiner, D.T. 1998. Soils in Our Environment, 8th ed. Prentice-Hall, Upper Saddle River, NJ, US.
  • Minchin, P. 1990. DECODA – Database for Ecological Community Data, Version 3.1. Research School of Pacific Studies, Canberra, ACT, AUS.
  • Packard, S. & Mutel, C.F. 2005. The Tallgrass Restoration Handbook: For Prairies, Savannas, and Woodlands, 2nd ed. Island Press, Washington, DC, US.
  • Partel, M. & Zobel, M. 1999. Small-scale plant species richness in calcareous grasslands determined by the species pool, community age and shoot density. Ecography 22: 153159.
  • Pickett, S.T.A. 1989. Space-for-time substitution as an alternative to long-term studies. In: Likens, G.E. (ed.) Long-Term Studies in Ecology: Approaches and Alternatives, pp. 110135. Springer, New York, NY, US.
  • Piper, J.K., Schmidt, E.S. & Janzen, A.J. 2007. Effects of species richness on resident and target species components in a prairie restoration. Restoration Ecology 15: 189198.
  • Polley, H.W., Derner, J.D. & Wilsey, B.J. 2005. Patterns of plant species diversity in remnant and restored tallgrass prairie. Restoration Ecology 13: 480487.
  • Potter, K.N., Torbert, H.A., Johnson, H.B. & Tischler, C.R. 1999. Carbon storage after long-term grass establishment on degraded soils. Soil Science 164: 718725.
  • Risser, P.G., Birney, E.C., Blocker, H.D., May, S.W., Parton, W.J. & Weins, J.A. 1981. The True Prairie Ecosystem. Hutchinson Ross Publishing Co., Stroudsburg, PA, US.
  • Ruiz-Jaen, M.C. & Aide, T.M. 2005. Restoration success: how is it being measured? Restoration Ecology 13: 569577.
  • Sala, O.E., Parton, W.J., Joyce, L.A. & Lauenroth, W.K. 1988. Primary production of the central grassland region of the United States. Ecology 69: 4045.
  • Samson, F. & Knopf, F. 1994. Prairie conservation in North America. BioScience 44: 418422.
  • Schramm, P. 1992. Prairie restoration: a twenty-five year perspective on establishment and management. In: Smith, D.D. & Jacobs, C.A. (eds.) Proceedings of the 12th North American Prairie Conference, pp. 169177. University of Northern Iowa, Cedar Falls, IA, US.
  • Seabloom, E.W. & Van Der Valk, A.G. 2003. The development of vegetative zonation patterns in restored prairie pothole wetlands. Journal of Applied Ecology 40: 92100.
  • Sheley, R.L. & Half, M.L. 2006. Enhancing native forb establishment and persistence using a rich seed mixture. Restoration Ecology 14: 627635.
  • Sivicek, V.A. & Taft, J.B. 2011. Functional group density as an index for assessing habitat quality in tallgrass prairie. Ecological Indicators 11: 12511258.
  • Sluis, W. 2002. Patterns of species richness and composition in re-created grassland. Restoration Ecology 10: 677684.
  • Smith, M.D. & Knapp, A.K. 1999. Exotic plant species in a C4 dominated grassland: invasibility, disturbance, and community structure. Oecologia 120: 605612.
  • Smith, R.S., Shiel, R.S., Bardgett, R.D., Millward, D., Corkhill, P., Rolph, G., Hobbs, P.J. & Peacock, S. 2003. Soil microbial community, fertility, vegetation and diversity as targets in the restoration management of a meadow grassland. Journal of Applied Ecology 40: 5164.
  • Sollins, P., Glassman, C., Paul, E.A., Swanston, C., Lajtha, K., Heil, J.W. & Elliot, E.T. 1999. Soil carbon and nitrogen. In: Robertson, G.P., Coleman, D.C., Bledsoe, C.S. & Sollins, P. (eds.) Standard Soil Methods for Long-Term Ecological Research, pp. 89105. Oxford University Press, New York, NY, US.
  • Spyreas, G., Meiners, S.J., Matthews, J.W. & Molano-Flores, B. 2012. Successional trends in floristic quality. Journal of Applied Ecology 49: 339348.
  • Swink, F. & Wilhelm, G. 1994. Plants of the Chicago Region, 4th ed. Indiana Academy of Science, Indianapolis, IN, US.
  • Taft, J.B., Wilhelm, G.S., Ladd, D.M. & Masters, L.A. 1997. Floristic quality assessment for vegetation in Illinois, a method for assessing vegetation integrity. Erigenia 15: 395.
  • Taft, J.B., Hauser, C. & Robertson, K.R. 2006. Estimating floristic integrity in tallgrass prairie. Biological Conservation 131: 4251.
  • Taylor, D.R., Aarssen, L.W. & Loehle, C. 1990. On the relationship between r/K selection and environmental carrying capacity: a new habitat template for plant life history strategies. Oikos 58: 239250.
  • Tilman, D. 1997. Community invisibility, recruitment limitation, and grassland biodiversity. Ecology 78: 8192.
  • Walker, L.R., Wardle, D.A., Bardgett, R.D. & Clarkson, B.D. 2010. The use of chronosequences in studies of ecological succession and soil development. Journal of Ecology 98: 725736.
  • Williams, D.W., Jackson, L.L. & Smith, D.D. 2007. Effects of frequent mowing on survival and persistence of forbs seeded into a species-poor grassland. Restoration Ecology 15: 2433.
  • Zedler, J.B. 2007. Success: an unclear, subjective descriptor of restoration outcomes. Ecological Restoration 25: 162168.
  • Zwicker, S.E. 1985. Soil Survey of Lee County, Illinois. National Cooperative Soil Survey. United States Department of Agriculture, Washington, DC, US.

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
avsc12051-sup-0001-AppendixS1.pdfapplication/PDF147KAppendix S1. Details of restoration site soil characteristics, plant composition and prescribed fire history.
avsc12051-sup-0004-AppendixS2.pdfapplication/PDF154KAppendix S2. Floristic quality assessment (FQA).
avsc12051-sup-0005-AppendixS3.pdfapplication/PDF163KAppendix S3. Methodology employed to examine relationships between Mean C, FQI, ANPP and total soil N as dependent variables and independent variables, and results of correlation and AIC model selection.
avsc12051-sup-0009-AppendixS4.pdfapplication/PDF115KAppendix S4. Chronosequence for species richness and Simpson's diversity index of the 19 prairie restorations.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.