Spatially modelling native vegetation condition

Summary   The assessment of vegetation condition is seen as an increasingly important requirement for effective biodiversity conservation in Australia. Condition assessments that operate at the scale of the site are well established. However, there is a need for mapped representations of vegetation condition at regional scales to: (i) assist with regional planning and target setting; (ii) provide regional context for site-based assessment; and (iii) monitor the change in vegetation condition at multiple scales. This paper describes a methodology for converting site condition data collected in plots into maps of vegetation condition across entire regions using a predictive statistical modelling framework (Generalized Additive Modelling) combined with a GIS. The research demonstrates how explanatory variables including topographic position, terrain roughness, landscape connectivity and remote sensing derived indices can be used to map the condition of native vegetation at the scale of a subcatchment. The inclusion of indices derived from remotely sensed imagery (SPOT4) as explanatory variables in the modelling is a novel component of this research. Although the methodology generates statistically and ecologically plausible models of vegetation condition, there are nevertheless limitations associated with the way plot data were collected and some of the explanatory variables, which impacts upon model utility. We discuss how these problems can be minimized when embarking upon studies of this type. We demonstrate how maps produced from exercises such as this could be used for conservation planning and discuss the limitations of these data for monitoring.     

Fauna habitat modelling and mapping: A review and case study in the Lower Hunter Central Coast region of NSW

Abstract Habitat models are now broadly used in conservation planning on public lands. If implemented correctly, habitat modelling is a transparent and repeatable technique for describing and mapping biodiversity values, and its application in peri‐urban and agricultural landscape planning is likely to expand rapidly. Conservation planning in such landscapes must be robust to the scrutiny that arises when biodiversity constraints are placed on developers and private landholders. A standardized modelling and model evaluation method based on widely accepted techniques will improve the robustness of conservation plans. We review current habitat modelling and model evaluation methods and provide a habitat modelling case study in the New South Wales central coast region that we hope will serve as a methodological template for conservation planners. We make recommendations on modelling methods that are appropriate when presence‐absence and presence‐only survey data are available and provide methodological details and a website with data and training material for modellers. Our aim is to provide practical guidelines that preserve methodological rigour and result in defendable habitat models and maps. The case study was undertaken in a rapidly developing area with substantial biodiversity values under urbanization pressure. Habitat maps for seven priority fauna species were developed using logistic regression models of species‐habitat relationships and a bootstrapping methodology was used to evaluate model predictions. The modelled species were the koala, tiger quoll, squirrel glider, yellow‐bellied glider, masked owl, powerful owl and sooty owl. Models ranked sites adequately in terms of habitat suitability and provided predictions of sufficient reliability for the purpose of identifying preliminary conservation priority areas. However, they are subject to multiple uncertainties and should not be viewed as a completely accurate representation of the distribution of species habitat. We recommend the use of model prediction in an adaptive framework whereby models are iteratively updated and refined as new data become available.     

Extended statistical approaches to modelling spatial pattern in biodiversity in northeast New South Wales. I. Species-level modelling

Statistical modelling of biological survey data in relation to remotely mapped environmental variables is a powerful technique for making more effective use of sparse data in regional conservation planning. Application of such modelling to planning in the northeast New South Wales (NSW) region of Australia represents one of the most extensive and longest running case studies of this approach anywhere in the world. Since the early 1980s, statistical modelling has been used to extrapolate distributions of over 2300 species of plants and animals, and a wide variety of higher-level communities and assemblages. These modelled distributions have played a pivotal role in a series of major land-use planning processes, culminating in extensive additions to the region's protected area system. This paper provides an overview of the analytical methodology used to model distributions of individual species in northeast NSW, including approaches to: (1) developing a basic integrated statistical and geographical information system (GIS) framework to facilitate automated fitting and extrapolation of species models; (2) extending this basic approach to incorporate consideration of spatial autocorrelation, land-cover mapping and expert knowledge; and (3) evaluating the performance of species modelling, both in terms of predictive accuracy and in terms of the effectiveness with which such models function as general surrogates for biodiversity.     

Modelling the distribution and compositional variation of plant communities at the continental scale

Aim

We investigate whether (1) environmental predictors allow to delineate the distribution of discrete community types at the continental scale and (2) how data completeness influences model generalization in relation to the compositional variation of the modelled entities.

Location

Europe.

Methods

We used comprehensive datasets of two community types of conservation concern in Europe: acidophilous beech forests and base‐rich fens. We computed community distribution models (CDMs) calibrated with environmental predictors to predict the occurrence of both community types, evaluating geographical transferability, interpolation and extrapolation under different scenarios of sampling bias. We used generalized dissimilarity modelling (GDM) to assess the role of geographical and environmental drivers in compositional variation within the predicted distributions.

Results

For the two community types, CDMs computed for the whole study area provided good performance when evaluated by random cross‐validation and external validation. Geographical transferability provided lower but relatively good performance, while model extrapolation performed poorly when compared with interpolation. Generalized dissimilarity modelling showed a predominant effect of geographical distance on compositional variation, complemented with the environmental predictors that also influenced habitat suitability.

Main conclusions

Correlative approaches typically used for modelling the distribution of individual species are also useful for delineating the potential area of occupancy of community types at the continental scale, when using consistent definitions of the modelled entity and high data completeness. The combination of CDMs with GDM further improves the understanding of diversity patterns of plant communities, providing spatially explicit information for mapping vegetation diversity and related habitat types at large scales.

     

Mapping vegetation in an arid,mountainous region of Western Australia

Abstract. Predictive mapping of vegetation using models linking vegetation units to mapped environmental variables has been advocated for remote areas. In this study, three different types of model were employed (within a GIS) to produce vegetation maps of the Hamersley Ranges region of Western Australia. The models were: (1) decision trees; (2) statistical models; and (3) heuristic/conceptual models. Maps were produced for three different levels of a floristic classification, i.e. 16 communities in two community groups with eight subgroups. All models satisfactorily established relationships between the vegetation units and available predictor variables, except where the number of sites of a particular unit was small. The different models often made similar predictions, especially for more widespread vegetation units. Map accuracy (as determined by field testing of maps) improved with increasing level of abstraction, with plant community maps ca. 50 % correct, subgroup maps ca. 60 % correct and group maps 90 % correct. Map inaccuracies were due to several factors, including low sample numbers producing unrepresentative models, poor resolution of and errors in available maps of predictor variables, and available predictor variables not being able to differentiate between certain vegetation units, particularly at the plant community level. Of these factors, poor resolution of maps was seen as the most critical. One type of model could not be recommended over another; however the choice of model will be largely dependent on the nature of the data set and the type of map coverage required.     

An evaluation of GIS-derived landscape diversity units to guide landscape-level mapping of natural communities

As conservation planning increases in scale from specific sites to entire regions, organisations like The Nature Conservancy face a critical need for GIS-based tools to evaluate landscapes on a regional scale. An existing, field-based approach to analyse the diversity of a landscape is by delineating natural community types, which is a time-intensive process. This study evaluated the utility of using an existing, GIS-derived landscape diversity model as a predictive tool for mapping natural communities on a large (8369 ha) upland forest site in the northern Taconic region of Vermont. The GIS model incorporates four geophysical factors: elevation, bedrock type, surficial deposits, and landform. A significant level (α=0.05) of association between eight pairs of landscape diversity unit (LDU) types and natural community types was found. However, the strength of these associations is low (Cramer's V values 0.172), suggesting a poor predictive efficiency of landscape diversity units for natural community types. The results suggest that variables in the LDU model are relevant to natural community distribution, but the LDU model alone is not an effective tool to aid in mapping of natural community types of upland forests in Vermont. Until better landscape-level techniques are developed, the role of this type of model is limited to screening the landscape for areas with a particular set of geophysical characteristics, which can help an ecologist interpret the patterns on the landscape, but cannot substitute for a field-based approach to natural community mapping.     

Potential natural vegetation: validity and applicability in landscape planning and nature conservation

Abstract. Since the introduction of ‘potential natural vegetation’ (PNV) as a concept in vegetation science by Tüxen (1956), many PNV-maps with different scales have been made. Tüxen emphasized the great value of PNV-maps for different purposes in land use, landscape planning and nature conservation, in particular with regard to forestry, agriculture and landscape management. Different aspects are discussed in order to examine the validity and applicability of PNV-maps in landscape planning and nature conservation. PNV-maps are useful for the differentiation of natural and landscape units on a small scale (< 1 : 100 000). However, maps of the potential natural vegetation are less useful for purposes of detailed planning on larger scales (> 1 : 100 000). Problems arise, for example, from the often highly hypothetical character of the construction and the practice of taking remnants of ‘natural’ vegetation as a reference object for the PNV. With regard to the goals of modern landscape planning and nature conservation purposes (e.g. conserving biodiversity in the cultural landscape of Central Europe) the exact documentation of the actual real vegetation (ARV) on intermediate and large scales gives much more detailed information than a hypothetical PNV.     

Vegetation mapping using hierarchical object‐based image analysis applied to aerial imagery and lidar data

Aims: The primary objective of this study is to map the distribution and quantify the cover of vegetation alliances over the entirety of San Clemente Island (SCI). To this end, we develop and evaluate the mapping method of hierarchical object‐based classification with a rule‐based expert system. Location: San Clemente Island, California, USA. Methods: We developed and tested an approach based on hierarchical object‐based classification with a rule‐based expert system to effectively map vegetation communities on SCI following the Manual of California Vegetation classification system. In this mapping approach, the shrub species defining each vegetation community and non‐shrub growth forms were first mapped using aerial imagery and lidar data, then used as input in an automated mapping rule set that incorporates the percent cover rules of a field‐based mapping rule set. Results: The final vegetation map portrays the distribution of 19 vegetation communities across SCI, with the largest areas comprised of California Annual and Perennial Grassland (35%) and three types of coastal sage scrub and maritime succulent scrub, comprising a combined 53% of the area. Map accuracy was assessed to be 79% based on fuzzy methods and 61% with a traditional accuracy assessment. The accuracy of tree identification was assessed to be 81%, but species‐level tree accuracy was 45%. Conclusions: Semi‐automated approaches to vegetation community mapping can produce repeatable maps over large spatial extents that facilitate ecological management efforts. However, some low‐statured shrub community types were difficult to differentiate due to patchy canopies of co‐occurring species including abundant non‐native grasses characteristic of complex disturbance histories. Species‐level tree mapping accuracy was low due to the difficulty of identifying species within poorly illuminated canyons, resulting from sub‐optimal image acquisition timing.     

Reliability of map accuracy assessments: A reply to Roff et al. (2016)

Roff et al. (Ecological Management and Restoration, 17 , 2016, 000) provide a discussion of the criteria expected for the best approach to validation of mapping programs and uses Hunter (Ecological Management & Restoration 17 , 2016, 40) to highlight issues involved. While we support the general principles outlined, we note that the review does not apply the same standards to Sivertsen et al. (Greater Hunter Native Vegetation Mapping Geodatabase Guide (Version 4.0). Office of Environment and Heritage, Department of the Premier and Cabinet, Sydney, Australia, 2011), the original document critiqued by Hunter (Ecological Management & Restoration 17 , 2016, 40). The Hunter (Ecological Management & Restoration 17 , 2016, 40) validation was based on a larger sample size, greater sampling within mapping units and greater representation of landscapes than Sivertsen et al. (Greater Hunter Native Vegetation Mapping Geodatabase Guide (Version 4.0). Office of Environment and Heritage, Department of the Premier and Cabinet, Sydney, Australia, 2011). Survey and validation sites being placed along public roads and lands are common to both the general Office of Environment and Heritage (OEH) and Hunter (Ecological Management & Restoration 17 , 2016, 40) validation methodologies. Thus, the criticisms of Roff et al. (Ecological Management and Restoration, 17 , 2016, 000) of the Hunter (Ecological Management & Restoration 17 , 2016, 40) approach apply equally, if not more, to Sivertsen et al. (Greater Hunter Native Vegetation Mapping Geodatabase Guide (Version 4.0). Office of Environment and Heritage, Department of the Premier and Cabinet, Sydney, Australia, 2011). We outline in the article how the Roff et al. (Ecological Management and Restoration, 17 , 2016, 000) critique was selective and in some cases incorrect in its analysis of issues presented in Hunter (Ecological Management & Restoration 17 , 2016, 40) and did not apply the same criteria to their own work. We conclude by discussing future directions for validating and mapping vegetation communities.     

Do plant communities exist? Evidence from scaling‐up local species‐area relations to the regional level

Abstract. One long tradition in ecology is that discrete communities exist, at least in the sense that there are areas of relatively uniform vegetation, with more rapid change in species composition between them. The alternative extreme view is the Self‐similarity concept – that similar community variation occurs at all spatial scales. We test between these two by calculating species‐area curves within areas of vegetation that are as uniform as can be found, and then extrapolating the within‐community variation to much larger areas, that will contain many ‘communities’. Using the Arrhenius species‐area model, the extrapolations are remarkably close to the observed number of species at the regional/country level. We conclude that the type of heterogeneity that occurs within ‘homogeneous’ communities is sufficient to explain species richness at much larger scales. Therefore, whilst we can speak of ‘communities’ for convenience, the variation that certainly exists at the ‘community’ level can be seen as only a larger‐scale manifestation of micro‐habitat variation.     

Standard survey designs drive bias in the mapping of upland swamp communities

Vegetation maps are critical biodiversity planning instruments, but the classification of vegetation for mapping can be strongly biased by survey design. Standardization of survey design across different vegetation types is therefore increasingly recommended for vegetation mapping programs. However, some vegetation types have complex small‐scale vegetation patterns that are important in characterizing these vegetation types, and standard designs will often not capture these patterns. The objective of this paper was to investigate the magnitude of potential map bias that results from survey design standardization and recommend approaches to deal with this bias. We surveyed upland swamps of the Greater Blue Mountains World Heritage Area Australia using two contrasting survey designs, including the standard 400 m² single quadrat design recommended and used by authorities. We then derived a classification for these swamps and tested the effect of survey design on this classification, species richness and the type of species detected (obligate or facultative swamp species). Species richness and species type were not significantly different among survey techniques. However, more than 40% of swamps clustered differently among survey designs. Thus, one of the 10 derived communities (which is floristically consistent with a previously mapped endangered community) was indistinct, and some individual swamps misclassified using the standard survey design. An effect of landscape position on swamp floristic patterns and a significant trend for high similarity scores among swamps surveyed with multiple small quadrats compared to the standard survey design was also determined. Australian upland swamps are classified at the global scale as shrub‐dominated wetlands, and complex floristic patterns have been recorded in shrub‐dominated wetlands in both northern and southern hemispheres. We therefore advocate either multiple survey designs or different survey standards for upland swamp communities and other vegetation types that have complex floristic patterns at small scales.     

Environmental Change and Human Health in Upper Hunter Communities of New South Wales,Australia

This article presents the theory and method informing an ongoing study of environmental change and human distress in the Upper Hunter Valley of New South Wales (NSW), Australia. The nature of environmental change in the Upper Hunter landscape over the past two centuries is first described, followed by the preliminary results of a long-term study that aims to investigate the nature of residents’ understanding of, and responses to, environmental change. Data from in-depth interviews found that the transformation of the environment from mining and power station activities was associated with significant expressions of distress linked to negative changes to interviewees’ sense of place, well-being, and control. A new concept, “solastalgia,” is introduced to help explain the relationship between ecosystem health, human health, and powerlessness. We claim that solastalgia, as opposed to nostalgia, is a type of homesickness (distress) that one gets when one is still “at home.” Future research will aim to validate a questionnaire to test the hypothesis that environmental distress is associated with levels of depression, quality of life, and rates of stress-related disease, as well as activism and environmental rehabilitation.     

Use of the ecological information system SynBioSys for the analysis of large datasets

The rapid developments in computer techniques and the availability of large datasets open new perspectives for vegetation analysis aiming at better understanding of the ecology and functioning of ecosystems and underlying mechanisms. Information systems prove to be helpful tools in this new field. Such information systems may integrate different biological levels, viz. species, community and landscape. They incorporate a GIS platform for the visualization of the various layers of information, enabling the analysis of patterns and processes which relate the individual levels. An example of a newly developed information system is SynBioSys Europe, an initiative of the European Vegetation Survey (EVS). For the individual levels of the system, specific sources are available, notably national and regional Turboveg databases for the community level and data from the recently published European Map of Natural Vegetation for the landscape level. The structure of the system and its underlying databases allow user‐defined queries. With regard to its application, such information systems may play a vital role in European nature planning, such as the implementation the EU‐program Natura 2000. To illustrate the scope and perspectives of the program, some examples from The Netherlands are presented. They are dealing with long‐term changes in grassland ecosystems, including shifts in distribution, floristic composition, and ecological indicator values.     

Associations between wasp communities and forest structure: Do strong local patterns hold across landscapes?

Abstract Forest structure and habitat complexity have been used extensively to predict the distribution and abundance of insect assemblages in forest ecosystems. We tested empirically derived predictions of strong, consistent relationships between wasp assemblages and habitat complexity, using both field assessments and vegetation indices from remote sensing as measures of habitat complexity. Wasp samples from 26 paired ‘high and low’ complexity sites in two forests approximately 70 km apart, were compared with normalized difference vegetation indices (NDVIs) derived from multispectral videography of the survey sites. We describe a strong unequivocal link between habitat complexity and wasp communities, the patterns holding over coarse and fine landscape scales. NDVIs were also excellent predictors of habitat complexity and hence wasp community patterns. Sites with greater NDVIs consistently supported a greater abundance and species richness, and a different composition of wasps to sites with low NDVIs. Using vegetation indices from remote sensing to gauge habitat complexity has significant potential for ecosystem modelling and rapid biodiversity assessment.     

Habitat selection of breeding riparian birds in an urban environment: untangling the relative importance of biophysical elements and spatial scale

Aim Urbanization is a leading threat to global biodiversity, yet little is known about how the spatial arrangement and composition of biophysical elements – buildings and vegetation – within a metropolitan area influence habitat selection. Here, we ask: what is the relative importance of the structure and composition of these elements on bird species across multiple spatial scales? Location The temperate metropolitan area of Cincinnati, Ohio, USA. Methods We surveyed breeding birds on 71 plots along an urban gradient. We modelled relative density for 48 bird species in relation to local woody vegetation composition and structure and to tree cover, grass cover and building density within 50–1000 m of each plot. We used an information‐theoretic approach to compare models and variables. Results At the proximate scale, native tree and understory stem frequency were the most important vegetation variables explaining bird distributions. Species’ responses to landscape biophysical features and spatial scales varied. Most native species responded positively to vegetation measures and negatively to building density. Models combining both local vegetation and landscape information represented best or competitive models for the majority of species, while models containing only local vegetation characteristics were rarely competitive. Smaller spatial scales (≤ 500 m) were most important for 36 species, and eight species had best models at larger scales (> 500 m); however, several species had competitive models across multiple scales. Main conclusions Habitat selection by birds within the urban matrix is the result of a combination of factors operating at both proximate and broader spatial scales. Efforts to manage and design urban areas to benefit native birds require both fine‐scale (e.g., individual landowners and landscape design) and larger landscape actions (e.g., regional comprehensive planning).     

Australia’s Stock Route Network: 2. Representation of fertile landscapes

The Stock Route Network (SRN) of New South Wales (NSW) and Queensland is a large‐scale system of predominantly roadside remnant vegetation, which was established in the 1800s to allow livestock to be moved. Proposed changes to the management of the SRN could result in some portions of it being sold to private landholders, or subjected to long‐term set‐stocking. This may have potentially negative impacts on some of the values of the SRN. One key feature of the SRN is that it covers low‐lying parts of the landscape, which are poorly protected by national parks. To quantify this, we specifically analysed a 41 million hectare portion of the SRN which transects the NSW ‘wheat‐sheep belt’, characterising its representation of woody vegetation cover and topography, and contrasting this with the National Reserve System. Our analysis revealed that 55% of stock routes occur in low‐lying valley portions of the landscape, compared with only 6% of the National Reserve System. The SRN supports a wide range of vegetation types and, unlike the National Reserve System, is not biased towards heavily forested areas. White Box‐Yellow Box‐Blakeley’s Red Gum woodland, which is listed as critically endangered by the Australian Government, was recorded in 803 (or 17.5%) of the 4575 stock routes in our data set. In contrast, only 10 of the 335 reserves within our spatial study region are known to support small occurrences of this community. Our findings suggest that the protection of the SRN and National Reserve System together may fulfil the ‘representation’ goal of systematic conservation planning far better than the National Reserve System on its own. Future research should quantify which stock routes in particular should receive priority for protection.     

Vegetation change and conservation status of Coastal Upland Swamps

Coastal Upland Swamp communities are characterized by high biodiversity and provide habitat for a range of threatened flora and fauna. In this research project, we are monitoring swamp vegetation dynamics over decadal timescales and relating observed changes to environmental factors. We have also modelled potential effects of climate change on swamp distributions. We found that swamp communities are spatially dynamic, both internally and in relation to the woodland matrix. Transitions between communities depended on initial states. In addition, these water‐dependent communities appeared highly sensitive to projected climate change and their ‘Endangered’ status makes their active management a high priority. Improved understanding of dynamics at the community and landscape scale facilitates horizon scanning and improves our capacity to plan effective management interventions now and under future management and climate change scenarios.     

Mountain summer farms in Røldal,western Norway –; vegetation classification and patterns in species turnover and richness

This paper discusses vegetation types and diversity patterns in relation to environment and land-use at summer farms, a characteristic cultural landscape in the Norwegian mountains. Floristic data (189 taxa) were collected in 130 4-m² sample plots within 12 summer farms in Røldal, western Norway. The study was designed to sample as fully as possible the range of floristic, environmental, and land-use conditions. Vegetation types delimited by two-way indicator species analysis were consistent with results from earlier phytosociological studies. Detrended correspondence analysis and canonical correspondence analysis show that rather than being distinct vegetation types, the major floristic variation is structured along a spatial gradient from summer farm to the surrounding heathland vegetation. Species richness (alpha diversity) was modelled against environmental variables by generalized linear modelling and compositional turnover (beta diversity) by canonical correspondence analysis. Most environmental factors made significant contributions, but the spatial distance-to-farm gradient was the best predictor of both species richness and turnover. While summer farms reduce mean species richness at the plot scale, the compositional heterogeneity of the upland landscapes is increased, thereby creating ‘ecological room’ for additional vegetation types and species. Within an overall similarity across scales, soil variables (pH, base saturation, LOI, phosphate and nitrogen) differed considerably in their explanatory power for richness and turnover. A difference between ‘productivity limiting’ factors and ‘environmental sieves’ is proposed as an explanation. Species turnover with altitude is relatively low in grasslands as compared to heaths.     

Zonal and azonal vegetation revisited: How is wetland vegetation distributed across different zonobiomes

In ‘zonal’ vegetation, climatic factors are the main influence on growth and performance and the climate determines the vegetation type completely, which makes this vegetation dominant in the landscape. If vegetation is ‘azonal’ however, local stresses are assumed to have an overwhelming influence on plant performance and climatic influences will be minimal; typically, this vegetation occurs only in small patches in the landscape. In this study I ask whether wetland plant communities, as they are described for South Africa, are evenly distributed among different terrestrial vegetation types, to test whether they are zonal or azonal. Three contingency tables were construed based on the counts of wetland vegetation records, defined on three hierarchical levels (Main Clusters, Community Groups and Community) and their occurrence in the country (at the level of Biome, Bioregion and terrestrial vegetation type). An ‘azonality index’ was calculated as the sum of all Chi‐square values for each wetland vegetation type divided by the total number of records. The overall correlation between hydroperiod and the azonality index was very weak. At the finest level, terrestrial vegetation types were clustered on the basis of having similar combinations of wetland community types. Eighteen different ‘wetland ecoregions’ have been defined, on the basis of wetland vegetation types occurring within them. Instead of regarding wetland vegetation as azonal, it should rather be regarded as ‘intrazonal’, meaning that climate does have an impact but many vegetation types are widespread across climatic regions. The reason why community types in wetlands are widespread is due to the monodominance of a single widespread, often clonal, species. The different wetland ecoregions do not correspond to terrestrial biomes, so it is expected that wetland vegetation responds differently to climate than terrestrial vegetation.