SEARCH

SEARCH BY CITATION

Keywords:

  • metric;
  • rapid biodiversity assessment;
  • scoring systems;
  • vegetation condition

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

Summary  Rapid, on-ground assessments of vegetation condition are frequently used as a basis for landholder education, development applications, distributing incentive funds, prescribing restoration treatments and monitoring change. We provide an overview of methods used to rapidly assess vegetation condition for these purposes. We encourage those developing new approaches to work through the steps we have presented here, namely define management objectives and operational constraints; develop an appropriate conceptual framework for the ecosystems under consideration; select an appropriate suite of indicators; and consider the options available for combining these into an index. We argue that information must be gained from broader scales to make decisions about the condition of individual sites. Remote sensing and spatial modelling might be more appropriate methods than on-ground assessments for obtaining this information. However, we believe that spatial prediction of vegetation condition will only add value to on-ground assessments rather than replace them. This is because the current techniques for spatially predicting vegetation condition cannot capture all of the information in a site assessment or at the required level of accuracy, and maps cannot replace the exchange of information between assessors and land managers that is an important component of on-ground assessment. There is scope for more sophistication in the way on-ground assessments of vegetation condition are undertaken, but the challenge will be to maintain the simplicity that makes rapid on-ground assessment a popular vehicle for informing natural resource management. We encourage greater peer review and publishing in this field to facilitate greater exchange of ideas and experiences.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

Rapid, on-ground assessments of vegetation condition for biodiversity at the scale of the stand, paddock or remnant – or site – are part of the daily routine of natural resource management in Australia. Rapid assessments based on easily measured biophysical attributes are frequently used as a basis for landholder education, development applications, supporting investment decisions (e.g. incentive funds), prescribing restoration treatments and monitoring change. Thus, assessments of this type are an important medium through which funding for, and regulation of, biodiversity are translated to on-ground actions.

Despite increasing popularity of the term, there is no standard definition of ‘vegetation condition’. It is a concept that reflects a desire to extend vegetation management from a concern about extent, type and configuration to one that also considers quality, health, function or viability. Vegetation condition is a value-laden concept that requires data to be interpreted through a ‘values prism’ along a continuum of ‘good’ to ‘bad’ (Fig. 1). For example, high perennial weed cover might constitute ‘bad’ condition in one context (e.g. conserving native plant species richness), but ‘good’ condition in another (e.g. lowering water tables for controlling salinity). Thus, when defining vegetation condition one must address questions such as: Good for what? Good for whom? Some definitions of vegetation condition are provided in Table 1.

image

Figure 1. Vegetation condition is an interpretation of biophysical data with respect to the values or context in which the assessment is undertaken (modified from Tongway & Ludwig 1997). For example, a high cover of exotic perennial grass species may be ‘poor’ condition in terms of native plant species composition, but ‘good’ condition in the context of salinity control.

Download figure to PowerPoint

Table 1.  Different definitions of vegetation condition, quality or health that have been used in Australia
Term usedDefinitionReference
Habitat qualityThe ability of key habitat components to supply the life requisites of selected species of wildlifeUS Fish and Wildlife Service (1980)
Rangeland healthThe sustained ability of land to produce forage from rainfallPickup et al. (1994)
Range conditionHas its own continuum . . . the position of a particular site along this continuum depends on a judgement of the value of the landscape for a given purposeTongway & Ludwig (1997)
ResilienceThe predicted degree to which the ecosystem retains a capacity to recover after the removal of the source problem and application of restoration treatmentsPerkins (2002)
Vegetation qualityThe degree to which the current vegetation differs from mature and apparently long-undisturbed stands of the same vegetation communityParkes et al. (2003)
Riparian conditionThe degree to which human-altered ecosystems diverge from local seminatural ecosystems in their ability to support a community of organisms and perform ecological functionsJansen et al. (2004)

There are numerous approaches employed to assess vegetation condition at the site level. In this paper, we provide an overview of these approaches in terms of the following:

  • • 
    Management or policy objectives and operational constraints
  • • 
    The conceptual framework used to describe the ecological system
  • • 
    Choice of vegetation attributes
  • • 
    Combining attributes into indices of condition

This information is presented to provide a framework of issues that should be considered when developing tools for rapidly assessing vegetation condition in the context of biodiversity. Some approaches for assessing vegetation condition include measures at broader scales (e.g. connectivity). We have confined our discussion to measures that can be made rapidly, on the ground, at the scale of the site that are indicative of current and future condition.

Management or policy objectives and operational constraints

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

Available expertise, time and resources and the objectives of the assessment have considerable bearing on the methods ultimately employed to assess vegetation condition. Management objectives range from those that are less rigid (e.g. sustainable land use) to more rigid (e.g. regulation) and resources can vary from relying on voluntary participation of non-professionals through to employment of full-time specialists. Examples of methods for assessing vegetation condition that are tailored to each combination of these broad objectives and operational constraints are provided in Table 2. These differ in terms of the type and number of biophysical attributes measured. For example, the tools developed for rapid use by non-specialists (e.g. Goldney & Wakefield 1997) generally require no species identification and are often based on recording the presence or absence of attributes, or assessing attributes in terms of broad abundance classes, rather than continuous measures. Tools intended for use by trained specialists (e.g. Holm et al. 1987) contain attributes that require considerable ecological expertise to measure. It is essential that the proposed methods are matched with the management objectives and the time, expertise and other resources available to the intended user.

Table 2.  Examples of vegetation condition assessment tools that have been produced in response to different management or policy objectives and different operational or resource constraints
  Management or policy objectives
Sustainable land useProtection or rehabilitationRegulation
Available time, resources and expertiseLowLandholder self-assessment (e.g. Goldney & Wakefield 1997)Incentive programs (e.g. Freudenberger 2001)Landholder self-assessment (e.g. Department of Water Land and Biodiversity Conservation 2002)
ModerateSite monitoring (e.g. Tongway & Hindley 2004)Incentive programs employing market-based instruments such as auctions (e.g. Parkes et al. 2003)Proposals to clear native vegetation (e.g. Gibbons et al. 2005)
HighLong-term monitoring of landscapes (e.g. Holm et al. 1987)Formal reservation (e.g. Margules & Pressey 2000)Environmental Impact Assessment (e.g. US Fish & Wildlife Service 1980)

The conceptual framework used to describe the ecological system

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

Toolkits or indices developed to assess vegetation condition represent predictive models of ecosystems (or components thereof) and therefore should be based on a considered conceptual framework that describes the ecological system. Different ecological frameworks or models that underpin vegetation condition assessment reflect the specific objectives of the different toolkits, but also inherent differences between ecosystems, different levels of understanding of ecological systems and different paradigms in ecology. Conceptual frameworks that underpin vegetation condition assessments include:

Succession

Succession is the orderly development of plant communities through a series of (seral) stages. Clements (1916, cited in Clements 1949) suggested that succession occurs deterministically towards a climax or stable state. Westoby et al. (1989) noted that successional theory predicts that disturbances (natural and unnatural) produce changes to vegetation condition in the opposite direction to climax and therefore can be reversed if the disturbance is removed. Assessment methods based on classical succession theory should therefore be avoided in ecosystems that do not revert to their previous condition after a disturbance is removed, or ecosystems that can develop into different states depending on the nature, or timing, of the disturbance. Assessment methods for old-growth forest (e.g. Resource and Conservation Assessment Council 1996), which is sometimes referred to as ‘late successional’ forest, is underpinned by successional theory.

State and transition

This conceptual framework reflects observations that the same ecosystem can occur in a range of alternative stable states (Westoby et al. 1989). Disturbance triggers transitions between states, but unlike classic successional theory, transitions do not necessarily occur in both directions or with equal ease. For example, a disturbance (e.g. cultivation) may move a system into an alternative stable state, but the removal of the disturbance may not result in the system returning to its previous state. Yates and Hobbs (1997) described the dynamics of temperate woodland ecosystems in agricultural landscapes in terms of this framework. Gibbons et al. (2005) draw on the state and transition framework when predicting the outcomes of management actions on vegetation condition.

Resilience

Resilience is the ‘degree, manner and pace of recovery after disturbance’ (Westman 1978). In more measurable terms it refers to the capacity of a community to recover (e.g. through resprouting, soil propagule banks and seed rain) after the removal of disturbances or stresses (sensu Westman 1978). The concept of resilience underpinned assessments of rangeland condition by Pickup et al. (1994) who measured the resilience of rangeland vegetation as forage production after rainfall – sites that produced most forage after rainfall had highest resilience. Resilience assessment also underpins condition class mapping widely used in Australia as a basis for prescribing restoration treatments and monitoring (e.g. Perkins 2002; Fig. 2).

image

Figure 2. An example of a decision tree developed in this case for assessing resilience/anticipated recovery capacity for cleared sites on the Cumberland Plain, NSW (summarized from Perkins 2002).

Download figure to PowerPoint

Trigger, transfer, reserve, pulse

This framework was originally developed by Ludwig and Tongway (1997) for vegetation in arid and semiarid environments. It is based on the notion that rainfall triggers biological, physical and chemical activities; the products of these activities are transferred across the landscape by water and wind, and deposited or reserved in different parts of the landscape and this produces a pulse in the vegetation (and other organisms). Using this approach sites are assessed in terms of their ‘leakiness’ (i.e. ability to conserve water and nutrients). Landscape Function Analysis (LFA) (Tongway & Hindley 2004) is an assessment toolkit based on these concepts. It is important to note that LFA is a measure of a site's ability to support perennial vegetation irrespective of the composition of that vegetation (e.g. whether exotic or native), so other measures are required if this is considered to be an important component of condition.

Reference conditions

A common approach that is used to assess vegetation condition is to compare the site of interest with sites of a comparable ecosystem type, but in relatively unmodified, nominally pristine or functional condition. The use of reference conditions is based on the premise that communities of biota are generally better adapted to, and an ecosystem as a whole functions better within, environments with relatively little contemporary anthropogenic modification (Landres et al. 1999), that an ecosystem is more resilient within its natural range of variation (Holling & Meffe 1996), or that ecosystems have intrinsic value and therefore restoration should strive to return them to their historic trajectory (Society for Ecological Restoration International Science and Policy Working Group 2004).

One criticism of this approach is that it could be misinterpreted as Clementsian (i.e. based on the notion of succession to climax), if the range of variation represented in the benchmark does not include the alternative states that an ecosystem may exhibit with environmental variation (Landres et al. 1999) or natural disturbance (McCarthy et al. 2004). Another criticism can be that, unless other information is used to inform management options for a site (e.g. the resilience of vegetation), then it assumes that each attribute has equal likelihood of restoration. The reference concept is useful if these potential pitfalls are addressed. Indeed, the reference concept has widespread use in rapid assessment across several countries (see reviews by Landres et al. 1999 and McElhinny et al. 2005) with several examples in Australia (e.g. Parsons et al. 2002; Parkes et al. 2003).

Choice of vegetation attributes

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

There are many biophysical attributes that can be measured as indicators of vegetation condition. Noss (1990) presented a useful framework for identifying indicators of vegetation condition based on three levels of organization within ecosystems: composition, structure and function. We have provided examples of attributes under each of these headings that are relevant in the context of biodiversity assessment at the scale of the site (Table 3). In their Rapid Appraisal of Riparian Condition (RARC) index, Jansen et al. (2004) presented a transparent approach for identifying attributes for use in a vegetation condition index. First, they listed the functions of vegetation (in the riparian zone) of interest (e.g. sediment trapping), then they listed the components of the riparian zone that performed these functions (e.g. ground cover), and finally they listed the indicators of these that were used in their index (e.g. leaf litter cover). We encourage the inclusion of functional indicators when assessing vegetation condition because the current condition of a site does not necessarily provide a complete picture of viability, resilience, or the trajectory of the site.

Table 3.  Potential attributes for assessing vegetation condition at the site grouped by composition, structure and function (sensu Noss 1990) (Sources: Noss 1990; Oliver 2002; Jansen et al. 2004; Perkins 2002; Tongway & Hindley 2004)
CompositionStructureFunction
• native plant species richness• cover by plant life form• presence of regeneration
• native plant species richness by life form• cover by vertical stratum• cover of bare ground
• cover of exotic species• number of vegetation strata• cryptogam cover
• presence/abundance of problematic weed species• tree diameter distribution• soil surface stability
• presence/abundance of threatened plant species• number of trees with hollows• rate of infiltration
• presence/abundance of increasers and/or decliners• volume (or other measure of abundance) of coarse woody debris• soil compaction
• presence/abundance of nectar or seed resources• adjacent land use
• mistletoe abundance• tree growth stage• dieback
• evidence of introduced animals (e.g. rabbits, foxes)• basal area of overstorey stems• soil salinity
• canopy height• presence/abundance of salt-tolerant plant species
• abundance of large, dead trees• presence/abundance of plant functional types
• litter cover (or other measure of abundance)• grazing, fire, or logging regime
• time since clearing
• rock cover• degree of soil modification
• mistletoe abundance
• perennial plant basal covera
• bioturbation

Attributes or indicators used to assess vegetation condition should meet the following criteria (Noss 1990; McElhinny et al. 2005):

  • • 
    Demonstrated ecological basis (i.e. significantly associated with the biota and processes of interest)
  • • 
    Applicable over the range of ecosystems and ecosystem states under consideration
  • • 
    Sufficiently sensitive to discriminate between the range of sites and states under consideration
  • • 
    Simple, cost-effective and repeatable to measure
  • • 
    Robust to seasonal or climatic variation
  • • 
    Instructive or helpful for assessors and managers with respect to interpreting and managing a site
  • • 
    Not highly correlated with other attributes being measured

Combining attributes into indices of condition

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

A common characteristic of methods developed to assess vegetation condition is that they combine data from multiple attributes into a simple index. Indices that combine multiple attributes have widespread use in society (e.g. All Ordinaries Index, Tertiary Entrance Rank) because they are conceptually easier to understand and compare than multiple attributes measured on different scales. There are several approaches typically used to combine raw vegetation attributes measured on sites into an index of vegetation condition – the more common ones are outlined below.

Using simple criteria to combine attributes

It may be appropriate to summarize unique combinations of attributes as discrete classes. For example, the method used to assess old-growth forest in New South Wales (Resource & Conservation Assessment Council 1996) combines two attributes of vegetation: percentage of senescent tree crowns in a stand, and the percentage of regrowth tree crowns in a stand. Each attribute is assigned to one of three levels (< 10%, 10–30%, > 30%), so a stand can be assigned to one of nine possible classes. This approach works because the nine classes are interpreted into one of only two possible outcomes (old-growth forest and not old-growth forest). However, using this system with just one additional attribute would produce 27 discrete, possible outcomes making it unsuitable for many applications.

Specific combinations of attributes can also be summarized as classes using stepwise approaches such as decision trees. Perkins (2002) developed a decision tree to assess the resilience or anticipated recovery capacity of sites on the Cumberland Plain in New South Wales (Fig. 2). This is a conceptually simple and transparent way of combining attributes that does not require any mathematical calculations or consideration of the extent to which an increase in one attribute compensates for the loss in another. However, the outcome is sensitive to the order in which the attributes appear – factors that appear earlier in the decision tree have greatest influence – and decision trees require continuous data to be summarized into categories (e.g. yes/no), thus potentially losing quantitative information (Williams & Araujo 2002).

Additive scoring systems

A common method used to combine multiple attributes into a metric, score or index for a site is to standardize the data for each attribute either: (i) as present or absent (0 or 1); (ii) as a score (e.g. 0, 1, 2 or 3); or (iii) by dividing the observed value by an expected value and summing the scores across all attributes. Attributes should be standardized relative to an appropriate benchmark (e.g. there should be a different overstorey cover benchmark in rainforest compared with woodland). The attributes may be subjectively weighted based on criteria such as their perceived importance for biodiversity or their ease of recruitment or replacement. Additive scoring systems underpin many rapid assessment protocols around the world such as the Index of Biotic Integrity (IBI) approach used by the US Environmental Protection Agency (Andreasen et al. 2001) and the Environmental Benefits Index that underpins the Conservation Reserve Program in the USA (Ribaudo et al. 2001). In Australia, examples include the Assessing Farm Bushland module in the Save the Bush Toolkit (Goldney & Wakefield 1997) (Table 4), Habitat Complexity Score (Catling & Burt 1995) (Table 5), Habitat Hectares (Parkes et al. 2003), the Biodiversity Benefits Index (Oliver & Parkes 2003) and the Rapid Appraisal of Riparian Condition (Jansen et al. 2004) (Table 6).

Table 4.  An example of a simple additive scoring system based on the presence/absence of features. This scoring system was developed for self-assessment of farm bushland in agricultural landscapes of central west New South Wales (Source: Goldney & Wakefield 1997)
QuestionTrue or false
  • *

    Scores of 16–20 are considered ‘quality bushland’, scores of 10–15 are considered ‘moderately degraded’ and scores < 10 are considered ‘highly degraded’.

1. Bushland is never or rarely grazed by stockFalse
2. Seedlings or suckers < 2 m tall are presentFalse
3. Bush is burned infrequently, i.e. at intervals > 5 yearsTrue
4. Trees mainly healthy, with little or no diebackTrue
5. There are no rabbitsFalse
6. Weeds are absent or found only around the edges.False
7. Trees are not harvested for timber or firewoodFalse
8. Free of land degradation, e.g. soil erosion, salinityTrue
9. Free of fertilizer application or herbicide driftTrue
10. < 20% of trees have mistletoeTrue
11. Free of vehicle tracks (other than fire trail)True
12. Free of feral predators such as cats and foxesFalse
13. Native shrubs are present in the understoreyFalse
14. Old-growth trees with hollows are presentTrue
15. Dead timber is left standingTrue
16. Fallen timber and logs are left on the groundFalse
17. Ground flora is mostly native grasses and herbsFalse
18. Ground is covered with litter of leaves, bark and twigsTrue
19. Bushland patch is > 0.6 haTrue
20. Joined to other bushland by bush corridor or scattered trees ≤ 50 m apartTrue
Total number of true answers11
Bushland health rating*Moderately degraded
Table 5.  A modified version of the Habitat Complexity Score (sensu Catling & Burt 1995) – an additive score used to assess woodland bird habitat (Watson et al. 2001)
AttributeScore = 0Score = 1Score = 2Score = 3Total
 0–10% cover10–20% cover20–50% cover> 50% cover 
Tree canopy   2
Tall shrub cover (2–4 m)   1
Short shrub cover (0.5–2 m)   0
 0–10% cover10–40% cover40–70% cover> 70% cover 
Ground herbage   1
Logs/rocks   1
Litter   2
Total Habitat Complexity Score    7
Table 6.  The Rapid Appraisal of Riparian Condition (RARC) (Jansen et al. 2004) – an example of an additive scoring system in which attributes are weighted differently
Sub-indexIndicatorScoreSub-index scoreWeighted score*
  • *

    Sub-index score divided by maximum subindex score and multiplied by 10.

HabitatWidth of riparian vegetation4/47/88.8
Longitudinal continuity of riparian vegetation (≥ 5 m wide)3/4  
CoverCanopy (> 5 m tall)2/35/124.2
Understorey (1–5 m tall)0/3  
Ground (< 1 m tall)2/3  
Number of layers1/3  
DebrisLeaf litter2/33/74.3
Standing dead trees (> 20 cm d.b.h.)0/1  
Fallen logs (> 10 cm diameter)1/3  
NativesCanopy (> 5 m tall)3/35/124.2
Understorey (1–5 m tall)0/3  
Ground (< 1 m tall)0/3  
Leaf litter2/3  
FeaturesNative canopy species regeneration1/21/71.4
Damage to regeneration0/2  
Native shrub/subcanopy regeneration0/2  
Reeds0/1  
Total (out of 50)   22.9

Additive scoring systems can result in some unwanted outcomes if not constructed carefully. For example, an increased contribution of one attribute can compensate for the loss in another. McCarthy et al. (2004) gave an example: the score lost from trees being cut down can be compensated by the increase in coarse woody debris. Weightings are often assigned to variables to overcome this issue, although in our experience this does not always overcome the problem. One option to overcome this problem is to multiply these attributes together rather than sum them.

Multiplicative scoring systems

Multiplying the values of attributes together recognizes that the condition of the site is defined by the co-occurrence of more than one attribute and the loss of one attribute cannot be directly compensated by the addition of another (Burgman et al. 2001). For example, suitable habitat for many species only occurs if feeding, shelter and nesting resources co-occur; or the viability of a site might be defined by the co-occurrence of mature plants and regeneration.

However, if a range of attributes are multiplied together then the assumption is that, if just one attribute is absent – and therefore gets a value of zero – then the site has no value. Although this can be true when considering habitat for a particular species (e.g. if nesting resources are absent then no amount of feeding resources will make the habitat suitable), it is not true when considering biodiversity in a broad sense as each attribute on its own generally affords some biodiversity value. McCarthy et al. (2004) proposed that a multiplicative approach works in the latter situation if the minimum value for each attribute is set to be greater than zero. Gibbons et al. (2005) addresses this issue by using a combination of addition and multiplication in their metric, which took the form:

  • a + b + c + (a × b)

where a, b, c are different attributes in the index and a, b are attributes that have more value where they co-occur. For example if a represented mature trees, b regeneration and c coarse woody debris, then some value would be given to a site with mature trees and coarse woody debris, but substantially more value to sites with mature trees, regeneration and coarse woody debris because such sites are likely to be more viable in the long term. Note that if using this approach then weightings used for variables in the additive part of the score may need to be different to weightings used for the interaction terms. One should also be aware that error can be magnified where variables are multiplied together.

Statistical approaches

Statistical methods represent a more objective alternative to scoring systems of the type discussed previously. For example, regression models can be developed to make predictions for a univariate response variable (e.g. the total number of species on a site) where a series of predictor variables are available. Models of this type lose their predictive ability outside the limits of the data used to build the model in the first place, so should not be used beyond this domain of application. Analyses from statistical models of this type are commonly used to inform which attributes are most appropriate, or the general way they are scored, rather than for assessing vegetation condition directly. Pattern-based statistics such as ordination and classification can be used to determine the difference between sites (e.g. a measured site and reference sites) simultaneously for many attributes. This difference, or measure of dissimilarity, then becomes the condition ‘score’. There are different measures of dissimilarity suitable for different types of data (e.g. Euclidian, Bray–Curtis) that are based on summing the differences between the values measured for each attribute on the sites that are to be compared. These approaches have been used to compare sites in terms of their species composition (e.g. Parsons et al. 2002), but can be used with suites of attributes of the type listed in Table 3.

Probability of persistence

The change in probability of persistence of species or communities across a region has been suggested as a currency that can be used for assessing the value of actions on individual sites (Williams & Araujo 2002; Faith et al. 2003). This approach evolves from thinking about ‘complementarity’, which is an estimate of the gain in biodiversity representation or persistence when a site is added to a network of areas managed for conservation (e.g. Margules & Pressey 2000). However, this approach can only be used where there are spatial data for each attribute across the entire region of interest so the change from adding, improving or removing attributes from sites can be calculated. Further, spatial datasets alone are typically too coarse for making predictions about an individual site, patch or paddock. However, there is considerable scope to link regional conservation assessment techniques with site assessment methods (S. Ferrier, pers. comm., 2005). This can be facilitated with spatial mapping of vegetation condition attributes.

Limitations of site-based vegetation condition assessments

The widespread use and influence of simple, rapid site-based assessments of vegetation condition indicates that they are an important vehicle for natural resource management. However, there are limitations of site-based assessments in the context of biodiversity conservation assessment that should be recognized.

Assessment methods of the type discussed here have very specific objectives or provide very generalized assessments of biodiversity. Such approaches are unlikely to capture all aspects of biodiversity (e.g. species with narrow habitat requirements). Further, it may not be appropriate to assess the priority of a site for conservation based on its existing condition alone, but also other factors such as the threats to a site. Thus, information on vegetation condition should only inform decision-making when used alongside other information.

Assessments of vegetation condition at individual sites should be informed by data on vegetation condition collected across the wider area. For example, Austin et al. (2000) reported that only 2% of sites dominated by Yellow Box (Eucalyptus melliodora Cunn. ex Schauer) in central west New South Wales had an understorey supporting greater than 50% cover of natives. Thus, highly modified examples of the Yellow Box community may be in good condition relative to the condition of the community as a whole. Further, several policies require vegetation condition to be reported or monitored at regional, state or national scales (e.g. Parkes & Lyon 2006). Although this information can be obtained by sampling landscapes using site-based assessments, it is unlikely that this would be the most efficient method given advances in spatial modelling and remote sensing that are presented elsewhere in this issue.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

Many approaches have been used to rapidly assess vegetation condition. We encourage those developing new approaches to work through the steps we have presented here, namely define management objectives and operational constraints; develop an appropriate conceptual framework for the ecosystems under consideration; select an appropriate suite of indicators; consider the options available for combining these into an index; and use data collected at broader scales to place the site in a landscape context.

There is not a strong culture of peer review and publication in this field, so there is not a large body of accessible literature in which the experiences from past projects can be readily accessed, examined and improved upon by others. All workers that have to develop methods for rapidly assessing vegetation condition are encouraged to document, seek peer review and publish their experiences to facilitate a greater interchange of ideas in this field.

Some methods for rapidly assessing vegetation condition at the scale of the site have been criticized within the scientific community because of their apparent simplicity. However, complex approaches can be difficult to understand or implement, which is an impediment to uptake by managers (Andreasen et al. 2001). Thus, complex methods for assessing vegetation condition might be more appropriate for testing and refining accepted systems, rather than representing the operational platform for vegetation condition assessment.

We have argued that information must be gained from broader scales to make decisions about the condition of individual sites. However, we believe that spatial representations of vegetation condition will only add value to site-based assessments rather than replace them. This is because the current techniques for mapping vegetation condition at broad scales cannot capture all of the information in a site assessment or at the required level of accuracy and maps cannot replace the exchange of information between assessors and landholders that is an important component of site-based assessment.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References

This paper benefited from comments by Danielle Ayers and Julian Seddon. The work was undertaken with financial contributions by the NSW Environment Trust and National Action Plan for Salinity and Water Quality. We thank two anonymous referees, Tein McDonald and Andre Zerger for providing comments that improved an earlier version of this manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Management or policy objectives and operational constraints
  5. The conceptual framework used to describe the ecological system
  6. Choice of vegetation attributes
  7. Combining attributes into indices of condition
  8. Conclusions
  9. Acknowledgements
  10. References
  • Andreasen J. K., O’Neill R. V., Noss R. and Slosser N. C. (2001) Considerations for the development of a terrestrial index of ecological integrity. Ecological Indicators 1, 2135.
  • Austin M. F., Cawsey E. M., Baker B. L., Yialeloglou M. M., Grice D. J. and Briggs S. V. (2000) Predicted Vegetation Cover in the Central Lachlan Region. CSIRO Sustainable Ecosystems, Canberra. Available from URL: http://www.cse.csiro.au/publications/2000/predictedvegetationcover.pdf.
  • Burgman M. A., Breininger D. R., Duncan B. W. and Ferson S. (2001) Setting reliability bounds on habitat suitability indices. Ecological Applications 11, 7078.
  • Catling P. C. and Burt R. J. (1995) Studies of the ground-dwelling mammals of eucalypt forests in south-eastern New South Wales: the effect of habitat variables on distribution and abundance. Wildlife Research 22, 271288.
  • Clements F. E. (1949) Dynamics of Vegetation: Selections from the Writings of Frederic E. Clements/compiled and edited by B. W. Allred and Edith S. Clements. H. W. Wilson, New York.
  • Department of Water, Land and Biodiversity Conservation (2002) Scattered tree habitat value ready reckoner – A guide for landholders. In: Scattered tree Clearance Assessment in South Australia (eds J. L.Cutten and M. W.Hodder), pp. 147166. Department of Water, Land and Biodiversity Conservation, Adelaide, SA.
  • Faith D. P., Carter G., Cassis G., Ferrier S. and Wilkie L. (2003) Complementarity, biodiversity viability analysis and policy-based algorithms for conservation. Environmental Science and Policy 6, 311328.
  • Freudenberger D. (2001) Bush for the Birds: Biodiversity Enhancement Guidelines for the Saltshaker Project, Boorowa, NSW. CSIRO Sustainable Ecosystems, Canberra. Available from URL: http://www.cse.csiro.au/publications/2001/Saltshaker.pdf.
  • Gibbons P., Ayers D., Seddon J., Doyle S. and Briggs S. (2005) BioMetric Version 1.8 A Terrestrial Biodiversity Assessment Tool for the NSW Property Vegetation Plan Developer Operational Manual. Department of Environment and Conservation (NSW), Canberra. Available from URL: http://www.nationalparks.nsw.gov.au/npws.nsf/content/biometric_tool.
  • Goldney D. and Wakefield S. (1997) Assessing farm bushland. In: Save the Bush Toolkit (Charles Sturt University and Orange Agricultural College) pp. 112. Charles Sturt University, Bathurst, NSW.
  • Holling C. S. and Meffe G. K. (1996) Command and control and the pathology of natural resource management. Conservation Biology 10, 328327.
  • Holm A. McR., Burnside D. G. and Mitchell A. A. (1987) The development of a system for monitoring trend in range condition in the arid shrublands of Western Australia. Australian Rangeland Journal 9, 1420.
  • Jansen A., Robertson A., Thompson L. and Wilson A. (2004) Development and application of a method for the rapid appraisal of riparian condition. River and Riparian Land Management Technical Guideline 4, 114. Available from URL: http://www.lwa.gov.au/downloads/publications_pdf/PR040656.pdf.
  • Landres P. B., Morgan P. and Swanson F. J. (1999) Overview of the use of natural variability concepts in managing ecological systems. Ecological Applications 9, 11791188.
  • Ludwig J. A. and Tongway D. J. (1997) A landscape approach to rangeland ecology. In: Landscape Ecology Function and Management. Principles from Australia's Rangelands (eds J.Ludwig, D.Tongway, D.Freudenberger, J.Noble and K.Hodgkinson), pp. 112. CSIRO Publishing, Collingwood, Vic.
  • Margules C. R. and Pressey R. L. (2000) Systematic conservation planning. Nature 415, 243253.
  • McCarthy M. A., Parris K. M., Van Der Ree R. et al. (2004) The habitat hectares approach to vegetation assessment: an evaluation and suggestions for improvement. Ecological Management & Restoration 5, 2427.
  • McElhinny C., Gibbons P., Brack C. and Bauhus J. (2005) Forest and woodland stand structural complexity: its definition and measurement. Forest Ecology and Management 218, 124.
  • Noss R. F. (1990) Indicators for monitoring biodiversity: A hierarchical approach. Conservation Biology 4, 355364.
  • Oliver I. (2002) An expert panel approach to the assessment of vegetation condition within the context of biodiversity conservation. Stage 1: The identification of condition indicators. Ecological Indicators 2, 223237.
  • Oliver I. and Parkes D. (2003) A Prototype Toolkit for Scoring the Biodiversity Benefits (and Disbenefits) of Landuse Change, Version 5. Department of Land and Water Conservation, Sydney.
  • Parkes D. and Lyon P. (2006) Towards a national approach to vegetation condition assessment that meets government investors’ needs: A policy perspective. Ecological Management & Restoration 7 (Suppl. 1), S3S5.
  • Parkes D., Newell G. and Cheal D. (2003) Assessing the quality of native vegetation: the ‘habitat hectares’ approach. Ecological Management & Restoration 4, S29S38.
  • Parsons M., Thoms M. and Norris R. (2002) Australian River Assessment System: AusRivAS Physical Assessment Protocol. Monitoring River Health Initiative Technical Report Number 22. Environment Australia, Canberra.
  • Perkins I. (2002) Harrington Park Stage 2 Ecological Assessment Final Report. Ian Perkins Consultancy Services and Aquila Ecological Surveys, Sydney.
  • Pickup G., Bastin G. N. and Chewings V. H. (1994) Remote sensing-based condition assessment for non-equilibrium rangelands under large-scale commercial grazing. Ecological Applications 4, 497517.
  • Resource and Conservation Assessment Council (1996) Draft Interim Forestry Assessment Report. New South Wales Government, Sydney.
  • Ribaudo M. O., Hoag D. L., Smith M. E. and Heimlich R. (2001) Environmental indices and the politics of the Conservation Reserve Program. Ecological Indicators 1, 1120.
  • Society for Ecological Restoration International Science and Policy Working Group (2004) The SER International Primer on Ecological Restoration. http://www.ser.org and Tucson, AZ: Society for Ecological Restoration International. Available from URL: http://www.ser.org/content/ecological_restoration_primer.asp.
  • Tongway D. and Hindley N. (2004) Landscape function analysis: a system for monitoring rangeland function. African Journal of Range and Forage Science 21, 109113.
  • Tongway D. J. and Ludwig J. A. (1997) The nature of landscape dysfunction in rangelands. In: Landscape Ecology Function and Management. Principles from Australian Rangelands (eds J.Ludwig, D.Tongway, D.Freudenberger, J.Noble and K.Hodgkinson), pp. 4962. CSIRO Publishing, Collingwood, Vic.
  • US Fish and Wildlife Service (1980) Habitat Evaluation Procedures. US Fish and Wildlife Service, Department of the Interior, Washington, DC. Available from URL: http://www.fws.gov/policy/ESMindex.html.
  • Watson J., Freudenberger D. and Paull D. (2001) An assessment of the focal-species approach for conserving birds in variegated landscapes in southeastern Australia. Conservation Biology 15, 13641373.
  • Westman W. E. (1978) Measuring the inertia and resilience of ecosystems. Bioscience 28, 705710.
  • Westoby M., Walker B. and Noy-Meir I. (1989) Opportunistic management for rangelands not at equilibrium. Journal of Range Management 42, 266274.
  • Williams P. H. and Araujo M. B. (2002) Apples, oranges and probabilities: Integrating multiple factors into biodiversity conservation with consistency. Environmental Modelling and Assessment 7, 139151.
  • Yates C. J. and Hobbs R. J. (1997) Woodland restoration in the Western Australian wheatbelt: A conceptual framework using a state and transitional model. Restoration Ecology 5, 2835.