You are on page 1of 11


Journal for Nature Conservation 14 (2006) 161—171

Monitoring ecological processes for
restoration projects
Jeffrey E. Herricka,, Gerald E. Schumanb, Albert Rangoa

USDA-ARS Jornada Experimental Range, MSC 3JER, NMSU, Box 30003, Las Cruces, NM 88003-8003, USA
USDA-ARS High Plains Grasslands Research Station, RRR 8408 Hildreth Road, Cheyenne, WY 82009, USA

Received 16 February 2006; accepted 28 April 2006

Ecosystem services;
Restoration of ecological processes is key to restoring the capacity of ecosystems to
support social, economic, cultural and aesthetic values. The sustainability of the
restored system also depends on processes associated with carbon, nutrient and
hydrologic cycles, yet most restoration monitoring is limited to plant community
composition. Our research has shown that short-term plant composition monitoring
Spatial pattern;
is a necessary but insufficient predictor of long-term restoration success. Long-term
Spatial variability
(up to 75 years) studies in the western United States show that short-term
monitoring of plant community composition alone incorrectly predicted the failure
of treatments that were ultimately successful, and the success of treatments that
ultimately failed. We propose that vegetation composition monitoring be combined
with one or more ecological process indicators reflecting changes in three
fundamental ecosystem attributes on which restoration success depends: soil and
site stability, hydrologic function and biotic integrity. These simple, rapid, plot-level
indicators reflect changes in resource redistribution and vegetation structure. We
include a case study involving restoration of mixed grass prairie on mineland in the
west-central United States.
Published by Elsevier GmbH.

Introduction time (Ruiz-Jaen & Aide, 2005). There are two
significant limitations to this approach. The first is
The definition of restoration success is commonly that it fails to reflect today’s diverse goals for
based on the presence, density, biomass or cover of ecological restoration, which increasingly include
one or more plant species at a particular point in the recovery and improvement of landscape
function to support multiple ecosystem services
Corresponding author. Tel.: +505 646 5194; (Aronson, Clewell, Blignaut, & Milton, 2006). In
fax: +505 646 5889. some cases, the plant community is restored, but
E-mail address: (J.E. Herrick). the ecosystem services demanded by society are

1617-1381/$ - see front matter Published by Elsevier GmbH.

cesses associated with soil organic matter accumu- Matson. in press. the dikes and changes in the fire regime and herbivore were abandoned as a restoration failure (Walton. Walton. 2002). depends on plant community composition and structure at multiple spatial scales. 2006). ARTICLE IN PRESS 162 J. The fire regime. These ‘‘slow variables’’ (Reynolds. Herrick. 1. a plant community-based approach subsequently established was clearly visible from leads to overpriced restoration efforts by requiring the air (Fig. Examples of restoration success following apparent Sosebee. Buonopane. Falk. as well as ignition sources. In 1979. ment in which annual precipitation averages jects emphasise the early stages of recovery 230 mm. much of which arrives as intense con- through site preparation necessary for plant estab. Huenneke. 1997. By 1997. the persistence of during the next four years. This is often the case perennial vegetation response to the slow rate of where sustaining high levels of agricultural produc. 2002. et al. of the ecosystem services valued by society. Perennial vegetation recovery 23 years after keystone ecological processes replace ‘‘reference 7. Cassman. In others.. Herrick. 2004) and Mediterranean steppes (Maestre & Cortina. 2006). the failure to re-establish one or more biophysical but the question of what needs to be managed is processes. & Herbel. southwestern United States. not. 2001) on similar soils in the particularly if these services are more effectively area. Sherwood. Havstad. & the establishment of late-successional or difficult Remmenga. function and ecological processes has been clearly documented in numerous other sys- tems. we propose that bare soil. The explain the subsequent loss of grasses and re- dikes were maintained. lation. light areas are Consistent with this proposal. including coastal sand dunes (Roze & Lemau- viel. & Havstad. This understanding has led to the proposal that ‘‘re- ference dynamics’’ based on an understanding of Figure 1. & Polasky. 2001). Gibbens.E. recovery of infiltration and nutrient cycling pro- tion is the primary objective (Tilman. re-establish desert grasslands on degraded soils cient predictor of long-term restoration success. periods favorable for grass processes on which the persistence of the restored establishment are relatively rare in this environ- plant communities depends. 2003) failure include the establishment of 7. Most restoration pro.. Walton (2005) attributes the lag in the provided by different species. biosolids were applied establishment of woody invasive species include once. For example. Dark areas are vegetated. following the resource islands associated with the woody species. Short-term monitoring of plant community compo. Naylor. sition alone incorrectly predicted the failure of The initial failure of grass establishment in arid and treatments that were ultimately successful and the semi-arid ecosystems is generally associated with success of treatments that ultimately failed. forest and savanna ecosystems often require reintroduction of a process such as fire (Allen et al. Gao & Reynolds. Process-related factors that may crusted loamy soil with a 1% slope in 1975.5 cm high soil berms were constructed in a vegetation. The impor- 2005). the perennial vegetation which tance of the relationship between vegetation structure. 2005). in turn. Valentin. In arid and semi-arid regions of the rarely explicitly addressed. the composition and to establish species early in the recovery process. vegetation composition indicators be supplemented .. 2001). in press). structure of the vegetation is similar to that of These species may not be essential to the recovery naturally occurring banded vegetation (Tongway.. Additionally.5 cm high which is related to both weather and ecohydrolo- runoff-barriers on a vegetation-free physically gical processes. & Rango. 2004). conditions’’ as the standard for restoration success free area in southern New Mexico (reprinted from Walton (Falk. inadequate soil moisture (Ethridge. populations (Herrick et al. apparent failure of the fourth seeding. many restoration projects that are adaptive management is increasingly required for touted as short-term successes do not persist due to long-term restoration success (Folke et al. Rango. Herrick et al. & Seghieri. Today. one-time shrub-re- Our research has shown that short-term plant moval and grass seeding are rarely sufficient to composition monitoring is a necessary but insuffi. 2004). (Herrick. and the area was seeded at least three times high rates of runoff and erosion. vective storms. It is widely acknowledged that active In contrast. 2001) The second limitation of using plant community appear to be limiting the recovery of this composition alone is that it ignores the ecological system. lishment.

restoration success. 2005b). particularly important for projects covering large Burkett. 2). areas where the current status and probable Figure 2. in turn. term monitoring indicators may also be used for itoring restoration success in arid and semi-arid short-term monitoring. Van Zee. monitoring rity. short-term monitoring. hydrologic function and biotic integ. ARTICLE IN PRESS Monitoring ecological processes 163 with one or more ecological process indicators Monitoring system reflecting changes in three fundamental ecosystem attributes on which restoration depends: soil and The ten-step. is completed prior to the accurately predict long-term restoration success initiation of the restoration project.and climate-defined ecological potential. status of key ecosystem attributes and processes cators across different ecosystems is provided in a (Step 3). beginning with the definition of restoration objec- acteristics to monitor changes in ecological status. Because these attributes module. is then used to also serve as the foundation for nearly all ecosystem document and. tion (Step 2) and the assessment of the current Guidance for flexibly interpreting individual indi. Indicators of vegetation Monitoring programme design should be an cover. composition and structure are interpreted integral part of the restoration planning process. based on developed iteratively with a landscape stratifica- its soil. the approach can also be used to may affect restoration success. The approach uses a combina- tion of soil and vegetation indicators that reflect the status of key ecological processes. The three basic measurements are used to generate indicators of the processes shown in the boxes which. are related to the three ecosystem attributes which serve as the foundation for most ecosystem services and are essential to the success of nearly all restoration projects. . iterative approach includes three site stability. tives and monitoring objectives (Step 1). together with soil surface and near-surface char. which are A unique reference is used for each site. 3). The first module. adjust factors that services (Fig. The landscape stratification (Step 2) is separate document (Herrick. & Whitford. basic modules (Fig. The third module help explain the importance of restoration projects involves more intensive long-term monitoring of to diverse groups in society. Some of the long- and illustrate a cost-effective approach for mon. This information is used to The specific objective of this paper is to present adjust management strategy. upland ecosystems. Havstad. where possible. A qualitative Monitoring programme design assessment protocol is used to select quantitative monitoring indicators. The second based on short-term data. This approach should allow managers to more programme design.

Each and transition models can be used to define what is ecological site may occur at multiple locations in possible and what is realistic. ecological states. topography and climate that has the within states. Current understanding of the status potential to support particular types and amounts of processes within states and factors associated of vegetation (Herrick. relatively easy. Herrick et al. associated text.. Krueger.. a unique undegraded state to a degraded state is also conceptual ‘‘state and transition model’’ (Bestel. 2005).. Volume II). like texture and depth. Restoration monitoring programme design and implementation (modified from Herrick et al. Stringham. altered to the point that simply removing a stress State and transition models consist of one or more is insufficient to promote recovery. Relatively reversible transitions occur similar Every 1-5 years: current and previous year's data.. Only relatively static soil proper.g. topography and climate (Step 2). The Six Steps measurements and measurement frequency Step 5: Select monitoring locations Step 6: Establish monitoring plots and record long-term monitoring data (baseline) Every year: Step 7: Record short-term monitoring data Short-term Step 8: Adjust management (e. Tugel. 2006). Interpret changes using short- term monitoring data Repeat Long-term Monitoring State and transition models used to interpret monitoring data relative to restoration objectives Step 10: Refine restoration objectives and strategy Figure 3. 2005b. grazing. Moving among plant a region based on soil and climate patterns. with transitions among states is described in the & Brown. Shifting from an on soil. An ecological site is a type of land with thresholds. Because these models are con- ties (Tugel et al. & Shaver. are ceptual. communities within a state is generally possible For each ecological site or strata defined based without significant external inputs. Moving from a degraded state to a meyer et al. state potential to support ecosystem services. treatment response are likely to vary spatially. Bestelmeyer. . they are easily updated by both scientists used to define ecological sites. irrigation or weed Monitoring control frequency) Step 9: Repeat long-term monitoring measurements. An ecological site and land managers.E. less degraded state requires significantly more 2003) is used together with assessment tools (Step 3) inputs because ecological processes have been to define realistic restoration objectives (Step 1). Step 1: Define/refine restoration and monitoring objectives Step 2:Stratify land into monitoring units (areas with similar characteristcs) Year 1: Step 3:Assess current status and develop/modify restoration Develop strategy Monitoring Program: Step 4: Select monitoring indicators. Archer. therefore includes areas with similar long-term Within the restoration planning process. ARTICLE IN PRESS 164 J. number of monitoring plots. 2003. The boundaries between states Ecological sites are widely applied in the United are marked by relatively irreversible transitions or States.

rills. & Herrick. Reference sheets are developed using all hydrologic function in areas where the ecological available sources of information. For example. and/or deposition areas S 7 Litter movement S 8 Soil surface resistance to erosion (soil stability in water test. blowouts. ARTICLE IN PRESS Monitoring ecological processes 165 A qualitative assessment protocol (Step 3). Similarly. . & Pellant. water 700 mm annual precipitation would be expected to flow patterns and large vegetation gaps are have short (less than 1 m) water flow patterns. associated with the undegraded or reference state ate the status of the land relative to its ecological (Herrick et al. Many of these improve hydrologic function.g. invasive species have mentation can help identify the processes that are negative effects on biotic integrity but sometimes likely to limit restoration success. Each of the three attributes is evaluated inde- Shaver.. Pyke. loamy soils on indicator of soil and site stability than it would be gently sloping terrain in a mid-latitude region with on the loamy soils described above. For example. (2005) for detailed explanation of each indicator. For ecological lands where the herbaceous component was lost. describing all 17 indicators for an ecological site is hydrologic function and biotic integrity (Pellant. sites with a state and transition model. is reference sheet describes the range of variability widely applied throughout North America to evalu. For example. set of processes are not masked by unrelated Applying this protocol prior to project imple.. or in shrub- tific literature and local knowledge. A typical reference sheet potential for three attributes: soil and site stability. infiltration and runoff) are very reasons that invasive species removals can have expensive to measure. 2001) 9 Soil surface loss or degradation S H B 10 Plant community composition and distribution relative to H infiltration and runoff 11 Compaction layer S H B 12 Functional/structural groups B 13 Plant mortality/decadence B 14 Litter amount H B 15 Annual production B 16 Invasive plants B 17 Reproductive capability of perennial plants B See Pellant et al. soil compaction. 2005. These would slow the movement of any runoff that did indicators would serve as indicators of degraded occur. Table 1. Herrick. ‘‘In. flow patterns. The protocol uses 17 easily pendently to ensure that serious problems in one observed soil and vegetation indicators (Table 1). approximately 1–2 pages long. the absence each indicator is developed for soils with similar of rill formation on a lakebed soil is a less important ecological potential. 2002). a reference sheet indicators to be variably weighted for different describing the range of variability expected for soils (Pyke et al. S H B Herrick et al. Ecological attribute assignments for 17 indicators used in the qualitative assessment protocol Indicator Attribute Soil & site Hydrologic Biotic integrity stability function 1 Rills S H 2 Water-flow patterns S H 3 Pedestals and/or terracettes S H 4 Bare ground S H 5 Gullies S H 6 Wind-scoured. but can be easily evaluated unintended consequences (Zavaleta. indicators. the terpreting Indicators of Rangeland Health’’. including scien. potential is for perennial grassland. Shaver. 2001). Pyke. 2006).. This is one of several processes (e. & using a combination of indicators including water Mooney. The soils should expected to dominate and their mere presence support relatively uniform vegetative cover that would have no effect on the evaluation. In ‘‘preponderance of evidence’’ approach allows order to increase consistency. no expected in areas where woody vegetation is rills and no compaction layer. and plant Another key feature of the protocol is that a community composition and spatial distribution. Hobbs. 2002).

2. (2005b).In undesirable at other stages in other ecosystems. locations within the area to be restored (Step 5) Herrick. However. see ‘‘Long-term monitoring’’) and cept. while ground effective. A stratified random sampling design is generally plant community composition and structure data to superior to randomly locating plots or transects allow vegetation-based objectives to be evaluated. both physical and effective as possible. while soil sheet (Herrick et al. been applied. gap intercept. & Whitford. indicators should reflect both the stated restora- off and are highly dispersible. such as floods. the of key indicators that reflect critical changes in the assessments are used to select monitoring indica. monitoring the factors which cover is the most important indicator of erosion contribute to restoration success (Step 7) can be resistance. each measurement should microbiotic crusts may be desirable at some stages generate multiple indicators that are relevant to of the restoration process in some ecosystems. because it focuses monitoring efforts on the most By quantifying the changes in both the plant important areas. In arid ecosystems. in fact. Unless the restoration protocol and others which generate one or more project has a research component. and vary with evaluate restoration success. Long-term monitoring any defined relationship between individual prop- erties.g. Consequently. For potential for wind erosion exists or vegetation example. processes required to achieve those objectives. climate and hydrology. The use of an ecological site-specific reference process. Guidance on the number of transects and The greatest interest in monitoring often exists measurements required to detect change at plot immediately after the restoration treatments have and landscape scales is provided in Herrick et al. Where the management to modify these factors (Step 8).E. in grazed systems. long-term monitoring increase water erosion because they increase run. Where fire is a critical completed on a subset of the points. Short-term monitoring All measurements are completed along trans- ects. The reference sheet approach is also distinguished from index-based approaches in that it does not assume a linear relationship or. Toledo (2004) . to adjust restoration soils. Litter cover is an important indicator of extremely valuable if it is possible to adjust decomposition and nutrient cycling. residual cover can be recorded to guide seasonal for wildlife habitat). Soil physical crusts objectives and to adapt the management strategy. height measurements can be grazing management. arid and semi-arid ecosystems. community and soil structure. plant canopy. however. com. In and may increase or decrease water infiltration order to make the monitoring system as cost- (Warren. While soil moisture monitoring if there is no way the indices are attractive (they are easily compared) manager can modify moisture availability. 2006) and the independent moisture should be measured or at least observed evaluation of three ecosystem attributes are two of in areas that can be irrigated or where evapotran- the critical differences between this assessment spiration can be modified. opportunities to apply local knowledge of the should be a part of any short-term monitoring relationships between ecosystem properties and programme. For example. basal cover and composition. and a soil stability test (Fig. 1995. and one or more restoration objectives and processes. These measurements also provide sufficient 6). an important indicator of production. it is indices that are independent of site potential (e. and can be quite effective where they have been documenting and being prepared to respond to rare calibrated. Van Zee. but To the extent possible.. irresponsible to recommend allocating resources to Tongway. Herrick et al. universal application can result in lost but potentially critical events. processes. processes and functions. Line-point intercept is used to quantify plete recovery takes decades and annual monitor. 2001). Burkett. forage utilisation and structure is an important restoration objective (i. limit wind erosion in low rainfall zones. where baseline measurements are completed (Step 2005a).e. ARTICLE IN PRESS 166 J. and ing of long-term indicators like perennial plant ground and plant litter cover. Microbiotic crusts tion objectives and the ecosystem properties and nearly always limit both wind and water erosion. fuel loads should be monitored. Finally. however. there are three In addition to identifying the processes that must basic measurements which together generate a set be addressed by the restoration treatments. they indirectly reflect nutrient cycling processes which are ex- pensive and difficult to monitor directly. The three primary objectives of repeating long- critical dynamic properties for limiting soil erosion term monitoring measurements (Step 9) are to are different for wind and water. three key ecosystem attributes: line-point inter- tors (Step 4. 2004). Havstad. Plant canopy cover is cover and soil structure is generally not cost. Tongway & Hindley.

Collins. patterns and processes varies across ecosys- together with the qualitative indicators used in the tems. which allows A number of user-friendly databases and field data 20 samples to be tested in 10 min. 2002). 2001) is used to rapidly Data that are not immediately summarised. and generate indicators that relate to hydrologic. the indicators can be used to develop Elzinga. advantage of the existing transect. 2005a. 1986. Blair. because the relationship between ecological prop- particularly when the measurements are interpreted erties.. such as the North American Great plant canopies or plant bases) longer than a Basin (Pyke & Knick. composition. How- measurements described in Herrick et al. & and precision increased. Briggs. By focusing only tion of these gaps is related to the ‘‘leakiness’’ of on those species of interest. Canopy cept. while more precise. relationship between structure. Breshears. USDA-ARS Jornada Experimental These three methods were selected because each Range. Additional recovery trajectories (Cortina et al. 2005b). Different processes are critical for the assessment process. Ludwig. Davenport. a belt transect can soil surface covered by functionally significant gaps be added. is sensitive to soil entry systems are now available. Wilcox.g. the cost of even these Steinauer. 2006) or microsite modification early detection of plant invasion. (2005b). This method. For example. This is much the methods provide a spatially explicit snapshot of more rapid than establishing formal plots. Bastin. & Herrick.. Eager. Another important selection functioning of tallgrass prairie ecosystems (fire criterion was measurement cost relative to informa. 2003). Careful review of restoration summer rainfall (runoff and erosion. The belt transect takes critical to restoration success because the distribu. measure soil surface and sub-surface macroaggre- analysed and interpreted are unlikely to be used. & Imeson. However. Tong. processes and alternative semi-quantitative methods that gener. The under canopy and interspace recovery can be pole is extended perpendicularly from the transect generated by combining line-point intercept data as necessary to determine whether or not an with gap intercept and soil stability data. Chewings. 2002. structure and stability. where trends in rare or exotic plant 2005b) is used to rapidly measure the proportion of species need to be monitored. In this modification of the con. 2006). the one or more measurements. 2005). which can This approach does not generate an index help elucidate the status of different processes. cover. Data analysis erosion and vegetation recovery processes. than plot-based methods. as illustrated by the case study below. 1998). individual falls within the transect. gate stability in water. 1998) than for arid and semi-arid simple measurements can exceed many restoration savanna ecosystems dominated by high-intensity monitoring budgets. & Resswell. Plot-based important processes. Wilcox. and monitoring objectives can be used to eliminate Breshears.. functions varies among systems. Also. Exotic plant species invasion is one of the between plants. Tongway . repeatable and versatile gaps are measured where wind erosion (Okin. Knapp. by walking along the transect with a pole of defined way. is notoriously poor for Gillette. systems (e. time costs are reduced the system (Ludwig. 1996) are ing relatively rare species of interest. A soil stability kit (Herrick et al. The line-point inter- minimum length (e. per 50-point transect were sufficient to generate Coughlan. The gap intercept method (Herrick et al. McKenzie. For example. and Gibbs (2001) and in ecosystem-specific indices. Many of these organic matter inputs and decomposition. measurements. 2006) automatically generate key indicators provides indicators that can be independently so that interpretations can be made as soon as data interpreted while yielding information that can be collection is complete. & Allen. 2005a. Willoughby. Plant basal gap intercepts are methods can be expensive due to the time required measured where increasing runoff resistance is for plot establishment. used to help interpret indicators provided by the other measurements. ARTICLE IN PRESS Monitoring ecological processes 167 found that height measurements at just 10 points soil science references (e. and may even be ally do not require transect establishment can be different during degradation and different types of applied (Herrick et al. Independent estimates of length: one-half the width of the desired plot.g. Where time is limited.g. Together. The spatially explicit data collection system Indices and additional indicators used for the first provides many opportunities to develop relationships among indicators. A ‘virtual plot’ is created Liedloff. ever. most important process in many arid and semi-arid tinuous line intercept method. can be used to monitor an indicator that was highly correlated with visual processes not specifically addressed by these three obstruction values measured with a cover pole. only gaps (between ecosystems. and for monitor- for seedling establishment (O’Connor. 30 cm) are measured.. Salzer. & tion provided. Klute. and grazing.

Litter cover was higher these observations. (2005a. We virtually identical at all three sites. indicating increasingly values these lands for other uses.. Average soil stability was also lower at both plant cover had been successfully restored at all reclaimed sites than at the native site. The results suggest that while measurements described above and used them to vegetation recovery is complete at Reclaimed 2.. application. however. Management practices. Based on are sufficient litter inputs. The predictive power of this 2–4 mm thick. soil surface fragments collected at approach can be enhanced by integrating it with random locations along the transects at each of the ecosystem. (Herrick et al. soil surface and sub-surface aggregate stability in Qualitative indicators. 6–8 mm diameter. indicated that randomly located 25 m transects at each of the runoff and erosion were occurring. For more detail on the methods used. USA At Mine 1. and associated negative feedbacks than at Native.g. 2005) suggested that while cover. much of the conceptual basis for the approach Soil stability was rated on a scale from 1 to 6 described here reflect Tongway’s earlier work (e. that increase intercanopy soil aggregate Line-point intercept was used to quantify the stability and reduce gap size by promoting perennial cover and composition of the dominant plant plant establishment could further reduce resource structural groups: shrubs. incomplete recovery. however.E. suggesting that observed differences relative to the reference sites soil structure recovers relatively rapidly where there in each of the three ecosystem attributes. Treatments were vegetation. Over 85% of the soil surface including recreation and watershed protection. Site characterisation data of native rangeland and reclaimed mine sites in northeastern Wyoming Mine 1 Mine 2 Native Reclaimed 1 Reclaimed 2 Native Reclaimed Elevation (m) 1521 1436 1458 1463 1486 Slope (%) 2–3 3–4 3–10 1–5 3–15 Recovery (years) n/a 4 10 n/a 10 Texture Sandy loam Fine sandy loam Loam Sandy loam Sandy loam . Soil stability three sites. Herrick et al. water. Table 2).and site-specific indicators of drivers five sites. Each measurement was completed along six water flow patterns and pedestalling. including litter movement. gap intercept. annuals. low rates. under perennial plant canopies. ecological processes were not function. Many of the indicators and plant canopies longer than 25 cm were recorded. perennial grasses and loss from these sites in the future. A rapidly surface exposed in large (450 cm) intercanopy gaps growing urban and exurban population. We also used the data to calculate native proach that has been successfully applied to a species canopy cover. such as mulch ence sites. Tongway. the traditional vegetation measure- In June 2005.. resulting in a higher proportion of the soil exclusively for livestock production. was ing at their full potential at two of the sites. All gaps between perennial number of ecosystems. 2006). The spatial distribution of the perennial (total of five sites. 2001) on 18. however. was exposed in 450 cm long intercanopy gaps at Indicators from our preliminary. but at relatively three reclaimed mine sites and two native refer. Reclaimed 1 due primarily to low perennial plant sessment (Pellant et al. qualitative as. we applied the long-term protocol ments indicated partial recovery at the more to evaluate restoration success at three sites recently restored site (Reclaimed 1) and near- relative to native reference areas on similar soils complete recovery at the older site (Reclaimed 2) at two surface coal mines in northeastern Wyoming (Table 3). and Hindley (2004) present an index-based ap. we selected the three basic under plant canopies. was more heterogeneous in implemented 4–10 years prior to measurement. Herrick et al. Case study: Mineland restoration in Results and discussion Wyoming. 2005b). Reclaimed 2 than in the native reference site This area has traditionally been used almost (Native). see (Herrick et al. ARTICLE IN PRESS 168 J. Table 2. 1995). in Reclaimed 2 (15%) than in Native (6%). the generate indicators that are relevant to each of the site remains more susceptible to soil erosion and attributes: line-point intercept. 2005b) and the dynamics of specific the ‘‘Long-term monitoring’’ section above and processes (Falk.

. which increased described here is that it ignores landscape level uniformity of microclimate and litter inputs. it is highly dispersed and et al. Since western societal values. At Mine 2 reclaimed site. evaluate restoration success relative to other um smithii) at the reclaimed site.8) 3.0(0. Vegetation cover and composition scape processes including soil erosion (Tongway & indicators suggested complete recovery at one of Hindley.2(0. At this reclaimed site. This appeared indicators suggest the opposite.0) 1.6(0. ARTICLE IN PRESS Monitoring ecological processes 169 Table 3. Animal activity and restoration . New indicators provided better protection from wind. Between-plant-canopy samples at the reclaimed site were similar to those at the native Future directions site. veloped and can be adapted to monitor restoration tion success showed incomplete recovery at the project success. 2005b).6) 2.6(0.9(0. especially in arid and semi-arid ecosys- tems where there is often a high level of resource redistribution among landscape units due to high rates of runoff. 2006).9(0. however.3(0.d. suggesting that significant recovery of soil structure had occurred.. and processes (Herrick et al. While not directly addressed in identical gap size distribution. Landscape-scale indicators are Mine 2 reclaimed site. This is largely due this case study. such as wildlife habitat (Herrick wheatgrass is rhizomatous.6) 2. and land managers was much more uniformly distributed than at the develop more effective restoration and post-re- Mine 1 reclaimed sites.0(0.7(0. and wind and water erosion. This may be due in part to The primary limitation of the monitoring approach the more dispersed plant spacing.6) n.7) 1. pattern with more mechanistic indicators of land- based indicators. Soil and vegetation indicators for the native and reclaimed sites in northeastern Wyoming (average7 standard error) Sample size (n) Mine 1 Mine 2 Native Reclaimed 1 Reclaimed 2 Native Reclaimed Canopy cover Total (%) 6 61(8) 51(5) 59(6) 71(7) 51(4) Native species (% of total plant vegetation) 6 100(0) 28(6) 94(3) 85(4) 70(4) Perennial grass (% of total vegetation) 6 80(6) 22(6) 83(5) 69(5) 70(2) Shrubs (% of total vegetation) 6 11(7) 1(1) 6(5) 9(1) 0(0) Soil surface in canopy gaps (%) 25–50 cm 6 14(3) 4(2) 18(1) 13(1) 14(2) 50–100 cm 6 6(4) 7(2) 9(3) 4(1) 7(3) 100–200 cm 6 0(0) 22(7) 4(2) 1(1) 2(2) 4200 cm 6 0(0) 59(11) 2(2) 0(0) 0(0) Soil stability (value from 1–6) Average 18 3. resulting in a virtually storation strategies. the more process-based necessary for both restoration project design and indicators showed near-complete recovery. regulatory agencies. These results reflecting landscape-scale processes are being de- suggest that while traditional indicators of restora.6) Under canopy 0–9a 2.b Intercanopy 9–18 3. the data can also be used to to the dominance of western wheatgrass (Pascopyr.2(0. does not form a significant canopy. resulting in fewer under canopy than intercanopy samples at most sites.8(0.7) 2.6) 3. Applying the two to be largely due to the absence of shrubs from the types of indicators together can help reclamation- site. the vegetation ists. and not yet recovered to levels recorded at the incomplete recovery at Mine 2.3) 3. These Summary and conclusions indicators will increasingly need to integrate land- scape ecology approaches that emphasise spatial The pilot study supports the need for process. As a result.6) a Samples were randomly located. 2004). The process-based corresponding native site (Table 3). monitoring.8) 1.3(0.7(0.7(0. no under canopy soil stability samples were collected at this site. b None of the sampling points fell completely beneath a plant canopy. total canopy cover had the reclamation sites at Mine 1 (Reclaimed 2).7) 2.7) 3.

2005b) should be identified erosion thresholds... N. shrubland and savanna ecosystems. Dave Pyke and Pat Shaver Remediation research at the Jornada: Past and future. J. Biology. tems: A broad perspective.. S. D.. supplementary methods and interpretation. J.. and Bernice Gamboa assisted with the manuscript Elmqvist... W. E. Vol. (2002). (2006). A. North Antelope/Rochell Mine. M. E. G. Deb Peters. & Herrick. Clewell. WY ing the effects of shifting rainfall seasonality and for their cooperation in locating reclaimed and event size over a landscape gradient. D. G.. 231–240. 2006. P.. R. and included in restoration monitoring programmes. T. D. Falk. Science. M. 280. 557–581. Herrick. Van Zee. C. Journal for NM: Distributed by University of Arizona Press. Matthias Gross.. 745–747.. Baeza. P. deSoyza. V. These dynamics are essential to Davenport. K. B. R. approach also fails to address episodic events. Sue Milton. (1998). R. W.. A. USDA-ARS Jornada Experimental Range. D. (1997). J. Burkett. Bestelmeyer. D. (2005a). Herrick. shrubland and savanna ecosystems. Seybold. 51. S. David Tongway. the wind and water vectors themselves (Peters. F.. in press. E. J. C. NM: Distributed by University of Arizona Swetnam. Pe´rez-Devesa.. Monitoring manual for 12. W. Vol. telmeyer. G. J. Kris Havstad.. Fernando Herrick. L.. Havstad.. Savage.. Monitoring plant and animal populations... E.. science-based arid land management. Maestre.. M. We also thank Jeremy Hayes.. Gunderson. E. A.. and two Brown. were also helpful in the development of this In: Structure and function of a Chihuahuan Desert manuscript. R. nature conservation and economics. J. L. Schulke. Tugel. transitions in the northern Chihuahuan Desert: Model- pany. 2006). (2003). Huenneke. H. & Gibbs. Historical shrub-grass Environmental Supervisor.. Oxford: Oxford University Press. An integrated framework for anonymous reviewers contributed useful sugges. Blignaut. Whitford. 1418–1433. R. & Reynolds. 35. Journal of Arid tions. and Systematics. A. (2001). M. Q. Herrick. M. Havstad. 9. & Milton. Chavez. the recovery of many systems (Falk. Briggs. Kennecott Energy Company. Carpenter. (2003).. Ethridge. Process-centered restoration. II: Aronson. & Allen.. Ed Fredrickson. J. (2006). R. and biodiversity in ecosystem management. Ecosystem struc- autonomously photograph restoration sites and ture. J. 56. M. C. T. J. A. as fire and flooding. 135–139. D. Jordi Cortina. S. T.E. Gillette. L. K.. M. E. Las Cruces. Global Change native rangeland sites on the mine permit areas. Walker. Powder River Coal Com.. C.. A.. restoration of southwestern ponderosa pine ecosys. Annual Review of Ecology Evolution mental Engineer. Ecological restoration: A new frontier for USDA-ARS Jornada Experimental Range.. M. Gao. Brown. Scheffer.. Havstad. J. D. in press). resolution remote sensing data which will increas- Cortina. such 14.. L. A. Journal of Range Management. Process. T. & press). Herrick. Environ. C.. 14. K.. Sosebee. 140–151. B. Design. E. This they related? Journal for Nature Conservation. 50. A. E. M. & Whitford. Havstad. W. G. B. Modulation of diversity by and processes will increasingly require ultra-high grazing and mowing in native tallgrass prairie. (in press). Archer. F. Allen. K. (2004)... I: Quick start. F. developed.. S. Salzer. K.. Jacobs Ranch Mine.. et al. et al... T. Vallejo... & Herbel. Knapp. Wilcox. et al. Van Zee.. grassland. & Rango. J. D.. W.. L. W. Alexander.. Episodic C. F. J. W. J. and restoration success: Are associated controls (Rango et al. J.. Joel Brown. T. W.. WY and Bryan Hansen. W. Van Zee. Malden. Regime preparation. Economic feasibility of range- Acknowledgments land seeding in the arid southwest. Mike Pellant. Sherwood.. G. Journal of Range Management. M.and ecosystem-specific monitoring must be Elzinga. J. 185–190. M. Ecological Applications.. M. D. D. 65. J. Cost-effective monitoring of these patterns Steinauer. Ecological Press. We thank Matt Mortenson for assistance with Falk. Breshears. MA: Blackwell Scientific. Don Falk. as well as current and potential future juniper ecosystems – a unifying perspective of soil drivers (Herrick et al. (1998). Development and use of state-and-transition models tions in resource redistribution processes and in both for rangelands.. Environments. (2006). K. A.. ingly be collected using aerial vehicles that can Valdecantos. M. K. Havstad. Ericha Courtright Folke.. Journal of Range Management. Gillette. Bestelmeyer. resilience. Discussions with Craig Allen. Journal data collection and Nicole Hansen for data entry for Nature Conservation. (2001). Soil aggregate stability kit for field-based soil . J.. Las Cruces. ecosystem: The Jornada Basin LTER. in Collins.. J. Herrick. 319–335. J. J. 114–126. Viewpoint: Sustainability of pinon- events. J. ecosystem function.. 1475–1493. shifts. and preliminary data analysis. Havstad.. & Whitford. M. L. treatments themselves can lead to further modifica. (2005b). 14. 152–160. Suckling. Monitoring manual for References grassland. Blair. Nature Conservation. & Schlesinger. Burkett.. ARTICLE IN PRESS 170 J. Willoughby. B. Herrick et al. Brandon Bes. & Maestre.

Spatial patterning of resource (2002). Shaver. (2005). P. N.. cess approach. function analysis: Procedures for monitoring and Multiscale controls on and consequences of aeolian assessing landscapes. MS Thesis. M.. 69. Environ.. & Schlesinger. Tongway.. 75–91. Laliberte. A. intensive production practices. (2005). J. B. soil Peters. P. Buonopane. Las evaluation. (in press). Herrick.. M. T. W. C.. D. Soil change. (1986). Australia: CSIRO processes in landscape change in arid and semiarid Sustainable Ecosystems. Sand dune restoration in American Society of Agronomy.. 15. K. J. Pyke. Persistence of biosolids in a Chihuahuan at proceedings of the IVth international rangeland Desert rangeland 18 years after application. Agricultural sustainability and degraded semiarid steppes. of Land Management. A... 149. D. D. Ludwig. & Herrick. Vegetation composition and structure McKenzie... Synopsis: Influence of biological soil & Havstad. & Liedloff. E. E. A. C. 1094–1106.. & Aide. Structure. 418. Nature. A. Tongway. J. M. in press.. T. (2003). 454–459. E. (2004).. Roanhorse. & Zavaleta.. (2005). 62. (1995). A. Gillette. Reynolds. V. 16. M. for management Ecological Studies. 584–597. 27–35. In Paper presented M. 2.). P. ARTICLE IN PRESS Monitoring ecological processes 171 quality and rangeland health evaluations. (2003).. S. Hierarchical control over seedling Tongway.). Landscape Okin. Shaver. Version 4. & Shaver. 17. & Hipple. C. 569–577. M. 494–502. In S... D. R. (2001). A. UAV utilization for rangelands: Viewing invasive species removal in a whole-ecosys- Current applications and future potentials. K. 37. (2006). success: How is it being measured? Restoration Landscape Ecology. CO: Bureau Tugel. Methods of soil analysis. Walton. Berlin: Springer. N... Part 1: Physical Academic Press. C. Eager.). J. (2004). CA: Klute. Biological soil crusts: Mexico. (2002). Soil as an indicator of rangeland wildlife habitat physical measurement and interpretation for land quality. A. A. E. Rango. Using historic data to assess crusts on arid land hydrology and soil stability. (Eds. environments: Ecological processes and consequences Interpreting indicators of rangeland health. P. T. E. Banded vegetation patterning in arid and semiarid Pellant.. (2001). Schmugge. 12.. Canberra.. Herrick. Interagency Technical Reference. (2006). North Brittany. G. change and landscape restoration. recruitment of the bunch-grass Themeda triandra in a Environmental Monitoring and Assessment. tem context. Journal of Applied Ecology. J.. Oxford University Press. Pyke. J. Coughlan. R. Huenneke. J. function and management (pp.. Puckett. D. 86.nmsu... & Hindley. (2001). Huenneke. N. W. Naylor. & Knick. S.. E. 56. (in press). C. Accessed 6 April 2006. Ph. Herrick. Ecology. & Cortina. W. in press. Valentin. J. J. Herrick. (2004). 157–171. Chewings. New Mexico State University. F. mental Practice. Herrick. M. L. 33. In Structure and function of a Chihuahuan land database and field data entry system. Encyclopedia of biodiversity. Warren.. M. A. NM.. B. K. J. Krueger. (1996). 55. Cruces. Journal of Arid Environments. Vegetation patches and State and transition modeling: An ecological pro- runoff-erosion as interacting ecohydrological pro. K.. (2005). Colingwood. R. F.. Jenkins. J. 253–275. D.. 44. Journal of Range Manage.. Restoration Ecology. Research and Management. F. E. A. Restoration Ecology. & ecosystem composition and function in a sequence of Polasky. A leakiness index for Ruiz-Jaen. T. Oxford: usda-ars... E. project in the Chihuahuan Desert.. global Walton. Mausbach. Insights into Tilman. 13. Levin (Ed. Plant invaders.. & Lemauviel. S.D. (2001). (2004). Vol. P. A. Australia: CSIRO Publishing. D. survey. Tongway. (2001). cesses in semiarid landscapes. & Remmenga. University of Dayton. T. Rango. J. (2005). & O. S. R. Stringham. interactive landscape model to solve old and new USDA-ARS Jornada Experimental Range. & Mooney. 223–232.. (2005). J. L. 671–677. A. Wilcox. and mineralogical methods (2nd ed... Restoration assessing landscape function using remote sensing. & Resswell. C.. G. L. CATENA. 29–35.. P.. and natural resources decision making: A L. O’Connor.. Range- problems. D. S. J. D. R. Breshears. Denver.php. 12. & Imeson. D. C.. effectiveness of shrub removal in southern New Belnap. J. J. Desertification. Journal of Range Management. 303–318. D. M. F.. ..). K. Dayton. & Herrick. H. Ludwig. P. Arid Land congress. & Pellant. J. M. Bastin. 288–297. H. Cassman.. 738–747. D. Ecology. D. G. 65. J. M. A. ment. D.. Havstad. V. Monitoring soil productive potential.. M. S. T. Journal of Arid Environments. Rangeland health attributes and indicators for accumulation in a 22 year-old water harvesting qualitative assessment. W. OH. (2002). Herrick.. Thesis. semi-arid savanna.. Maestre. Trends in Ecology and Evolution. USA: Roze. (2002). Gibbens. Hobbs. ing. http:// Desert ecosystem: The Jornada Basin LTER. & Seghieri. Soil Science Society of America in Jornada research: Developing and applying an Journal. Madison. San Diego. A. G. Brown. Pyke. Matson. Toledo. environments. H. Bestelmeyer. 349–360). France: A 10-year monitoring study.. Lange (Eds. J. L.. K. In J. 106–113. Future directions blueprint for action.. A.