A Robust Technique for Mapping Vegetation Condition Across a Major River System

Shaun Cunningham, Ralph MAC NALLY, Jennifer Read, Patrick Baker, Matt White, Jim THOMSON, P. Griffioen

    Research output: Contribution to journalArticle

    58 Citations (Scopus)

    Abstract

    Ecologists need to develop tools that allow characterization of vegetation condition over scales that are pertinent to species’ persistence and appropriate for management actions. Our study shows that stand condition can be mapped accurately over the floodplain of a major river system (ca 100,000 ha of forest over 1600 km of river)—the Murray River in southeastern Australia. It demonstrates the value of using quantitative ground surveys in conjunction with remotely sensed data to model vegetation condition over very large spatial domains. A comparison of four modelling methods found that stand condition was best modelled using the multivariate adaptive regression spline (MARS) method (R2 = 0.85), although there was little difference among the methods (R2 = 0.77–0.85). However, a subsequent validation survey of condition at new locations showed that use of artificial neural networks had substantially higher predictive power (R2 = 0.78) than the MARS model (R2 = 0.28). This discrepancy demonstrates the value of using several modelling approaches to determine relationships among vegetation responses and environmental variables, and stresses the importance of validating ecological models with predictive surveys conducted after model building. The artificial neural network was used to produce a stand condition map for the whole floodplain, which predicted that only 30% of the area containing Eucalyptus camaldulensis stands is currently in good condition. There is a downstream decline in stand condition, which is related to more extreme declines in flooding, due to water harvesting, and drier climate found in the Lower Murray region. Rigorous surveying and modelling approaches, such as those used here, are necessary if vegetation health is to be effectively monitored and managed.
    Original languageEnglish
    Pages (from-to)207-219
    Number of pages13
    JournalEcosystems
    Volume12
    DOIs
    Publication statusPublished - 2009

    Fingerprint

    vegetation mapping
    river system
    Rivers
    vegetation
    rivers
    Splines
    artificial neural network
    floodplain
    modeling
    Neural networks
    Surveying
    floodplains
    neural networks
    methodology
    river
    surveying
    persistence
    water harvesting
    flooding
    Health

    Cite this

    Cunningham, S., MAC NALLY, R., Read, J., Baker, P., White, M., THOMSON, J., & Griffioen, P. (2009). A Robust Technique for Mapping Vegetation Condition Across a Major River System. Ecosystems, 12, 207-219. https://doi.org/10.1007/s10021-008-9218-0
    Cunningham, Shaun ; MAC NALLY, Ralph ; Read, Jennifer ; Baker, Patrick ; White, Matt ; THOMSON, Jim ; Griffioen, P. / A Robust Technique for Mapping Vegetation Condition Across a Major River System. In: Ecosystems. 2009 ; Vol. 12. pp. 207-219.
    @article{2a349a56f08142b89532993ec204e0fe,
    title = "A Robust Technique for Mapping Vegetation Condition Across a Major River System",
    abstract = "Ecologists need to develop tools that allow characterization of vegetation condition over scales that are pertinent to species’ persistence and appropriate for management actions. Our study shows that stand condition can be mapped accurately over the floodplain of a major river system (ca 100,000 ha of forest over 1600 km of river)—the Murray River in southeastern Australia. It demonstrates the value of using quantitative ground surveys in conjunction with remotely sensed data to model vegetation condition over very large spatial domains. A comparison of four modelling methods found that stand condition was best modelled using the multivariate adaptive regression spline (MARS) method (R2 = 0.85), although there was little difference among the methods (R2 = 0.77–0.85). However, a subsequent validation survey of condition at new locations showed that use of artificial neural networks had substantially higher predictive power (R2 = 0.78) than the MARS model (R2 = 0.28). This discrepancy demonstrates the value of using several modelling approaches to determine relationships among vegetation responses and environmental variables, and stresses the importance of validating ecological models with predictive surveys conducted after model building. The artificial neural network was used to produce a stand condition map for the whole floodplain, which predicted that only 30{\%} of the area containing Eucalyptus camaldulensis stands is currently in good condition. There is a downstream decline in stand condition, which is related to more extreme declines in flooding, due to water harvesting, and drier climate found in the Lower Murray region. Rigorous surveying and modelling approaches, such as those used here, are necessary if vegetation health is to be effectively monitored and managed.",
    keywords = "Eucalyptus camaldulensis, floodplains, neural networks, regression splines, regression trees, remote sensing, river regulation, validation, vegetation condition.",
    author = "Shaun Cunningham and {MAC NALLY}, Ralph and Jennifer Read and Patrick Baker and Matt White and Jim THOMSON and P. Griffioen",
    year = "2009",
    doi = "10.1007/s10021-008-9218-0",
    language = "English",
    volume = "12",
    pages = "207--219",
    journal = "Ecosystems",
    issn = "1432-9840",
    publisher = "Springer",

    }

    Cunningham, S, MAC NALLY, R, Read, J, Baker, P, White, M, THOMSON, J & Griffioen, P 2009, 'A Robust Technique for Mapping Vegetation Condition Across a Major River System', Ecosystems, vol. 12, pp. 207-219. https://doi.org/10.1007/s10021-008-9218-0

    A Robust Technique for Mapping Vegetation Condition Across a Major River System. / Cunningham, Shaun; MAC NALLY, Ralph; Read, Jennifer; Baker, Patrick; White, Matt; THOMSON, Jim; Griffioen, P.

    In: Ecosystems, Vol. 12, 2009, p. 207-219.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - A Robust Technique for Mapping Vegetation Condition Across a Major River System

    AU - Cunningham, Shaun

    AU - MAC NALLY, Ralph

    AU - Read, Jennifer

    AU - Baker, Patrick

    AU - White, Matt

    AU - THOMSON, Jim

    AU - Griffioen, P.

    PY - 2009

    Y1 - 2009

    N2 - Ecologists need to develop tools that allow characterization of vegetation condition over scales that are pertinent to species’ persistence and appropriate for management actions. Our study shows that stand condition can be mapped accurately over the floodplain of a major river system (ca 100,000 ha of forest over 1600 km of river)—the Murray River in southeastern Australia. It demonstrates the value of using quantitative ground surveys in conjunction with remotely sensed data to model vegetation condition over very large spatial domains. A comparison of four modelling methods found that stand condition was best modelled using the multivariate adaptive regression spline (MARS) method (R2 = 0.85), although there was little difference among the methods (R2 = 0.77–0.85). However, a subsequent validation survey of condition at new locations showed that use of artificial neural networks had substantially higher predictive power (R2 = 0.78) than the MARS model (R2 = 0.28). This discrepancy demonstrates the value of using several modelling approaches to determine relationships among vegetation responses and environmental variables, and stresses the importance of validating ecological models with predictive surveys conducted after model building. The artificial neural network was used to produce a stand condition map for the whole floodplain, which predicted that only 30% of the area containing Eucalyptus camaldulensis stands is currently in good condition. There is a downstream decline in stand condition, which is related to more extreme declines in flooding, due to water harvesting, and drier climate found in the Lower Murray region. Rigorous surveying and modelling approaches, such as those used here, are necessary if vegetation health is to be effectively monitored and managed.

    AB - Ecologists need to develop tools that allow characterization of vegetation condition over scales that are pertinent to species’ persistence and appropriate for management actions. Our study shows that stand condition can be mapped accurately over the floodplain of a major river system (ca 100,000 ha of forest over 1600 km of river)—the Murray River in southeastern Australia. It demonstrates the value of using quantitative ground surveys in conjunction with remotely sensed data to model vegetation condition over very large spatial domains. A comparison of four modelling methods found that stand condition was best modelled using the multivariate adaptive regression spline (MARS) method (R2 = 0.85), although there was little difference among the methods (R2 = 0.77–0.85). However, a subsequent validation survey of condition at new locations showed that use of artificial neural networks had substantially higher predictive power (R2 = 0.78) than the MARS model (R2 = 0.28). This discrepancy demonstrates the value of using several modelling approaches to determine relationships among vegetation responses and environmental variables, and stresses the importance of validating ecological models with predictive surveys conducted after model building. The artificial neural network was used to produce a stand condition map for the whole floodplain, which predicted that only 30% of the area containing Eucalyptus camaldulensis stands is currently in good condition. There is a downstream decline in stand condition, which is related to more extreme declines in flooding, due to water harvesting, and drier climate found in the Lower Murray region. Rigorous surveying and modelling approaches, such as those used here, are necessary if vegetation health is to be effectively monitored and managed.

    KW - Eucalyptus camaldulensis

    KW - floodplains

    KW - neural networks

    KW - regression splines

    KW - regression trees

    KW - remote sensing

    KW - river regulation

    KW - validation

    KW - vegetation condition.

    U2 - 10.1007/s10021-008-9218-0

    DO - 10.1007/s10021-008-9218-0

    M3 - Article

    VL - 12

    SP - 207

    EP - 219

    JO - Ecosystems

    JF - Ecosystems

    SN - 1432-9840

    ER -