To what are woodland birds responding? Inference on relative importance of in-site habitat variables using several ensemble habitat modelling techniques

Jian D. L. Yen, Jim THOMSON, Peter Vesk, Ralph MAC NALLY

    Research output: Contribution to journalArticle

    16 Citations (Scopus)

    Abstract

    We sought to identify those in-site habitat characteristics that best predict distributions of woodland birds in the box ironbark region of central Victoria, Australia. Our focus was on comparing and melding outcomes from several forms of ensemble modelling methods, which account for uncertainty in model structure and allow assessments of variable importance. We used boosted regression trees (BRT), Bayesian additive regression trees (BART) and random forests (RF) to model bird occurrences for 47 species using 43 predictor variables measured at 184 2-ha sites. The majority of predictor variables were in-site habitat variables, but vegetation cover in the surrounding landscape (500 m radius) and geographic coordinates were included to account for known effects of habitat fragmentation and of geographic clines. A consensus model also was developed, built from averaged predictions from the three techniques. We subdivided the avifauna into guilds and other categories (e.g. conservation status) to examine whether there were differences among such subdivisions. Based on cross validation, the consensus model and RF performed best, followed by BART and then BRT. Of the in-site habitat variables, the basal area of red-ironbark trees and groundstorey characteristics such as fineand coarse-litter cover and litter depth had greatest influence on bird occurrences. These results can inform on-site restoration actions (what to restore) and, therefore, complement strategic landscape planning (where and when to restore).
    Original languageEnglish
    Pages (from-to)946-954
    Number of pages9
    JournalEcography
    Volume34
    DOIs
    Publication statusPublished - 2011

    Fingerprint

    woodlands
    woodland
    bird
    birds
    habitat
    habitats
    modeling
    Eucalyptus sideroxylon
    litter
    Victoria (Australia)
    model uncertainty
    landscape planning
    avifauna
    cline
    conservation status
    habitat fragmentation
    guild
    methodology
    vegetation cover
    basal area

    Cite this

    @article{094e417857764314945dfbc71443cb56,
    title = "To what are woodland birds responding? Inference on relative importance of in-site habitat variables using several ensemble habitat modelling techniques",
    abstract = "We sought to identify those in-site habitat characteristics that best predict distributions of woodland birds in the box ironbark region of central Victoria, Australia. Our focus was on comparing and melding outcomes from several forms of ensemble modelling methods, which account for uncertainty in model structure and allow assessments of variable importance. We used boosted regression trees (BRT), Bayesian additive regression trees (BART) and random forests (RF) to model bird occurrences for 47 species using 43 predictor variables measured at 184 2-ha sites. The majority of predictor variables were in-site habitat variables, but vegetation cover in the surrounding landscape (500 m radius) and geographic coordinates were included to account for known effects of habitat fragmentation and of geographic clines. A consensus model also was developed, built from averaged predictions from the three techniques. We subdivided the avifauna into guilds and other categories (e.g. conservation status) to examine whether there were differences among such subdivisions. Based on cross validation, the consensus model and RF performed best, followed by BART and then BRT. Of the in-site habitat variables, the basal area of red-ironbark trees and groundstorey characteristics such as fineand coarse-litter cover and litter depth had greatest influence on bird occurrences. These results can inform on-site restoration actions (what to restore) and, therefore, complement strategic landscape planning (where and when to restore).",
    author = "Yen, {Jian D. L.} and Jim THOMSON and Peter Vesk and {MAC NALLY}, Ralph",
    year = "2011",
    doi = "10.1111/j.1600-0587.2011.06651.x",
    language = "English",
    volume = "34",
    pages = "946--954",
    journal = "Ecography",
    issn = "0906-7590",
    publisher = "Wiley-Blackwell",

    }

    To what are woodland birds responding? Inference on relative importance of in-site habitat variables using several ensemble habitat modelling techniques. / Yen, Jian D. L.; THOMSON, Jim; Vesk, Peter; MAC NALLY, Ralph.

    In: Ecography, Vol. 34, 2011, p. 946-954.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - To what are woodland birds responding? Inference on relative importance of in-site habitat variables using several ensemble habitat modelling techniques

    AU - Yen, Jian D. L.

    AU - THOMSON, Jim

    AU - Vesk, Peter

    AU - MAC NALLY, Ralph

    PY - 2011

    Y1 - 2011

    N2 - We sought to identify those in-site habitat characteristics that best predict distributions of woodland birds in the box ironbark region of central Victoria, Australia. Our focus was on comparing and melding outcomes from several forms of ensemble modelling methods, which account for uncertainty in model structure and allow assessments of variable importance. We used boosted regression trees (BRT), Bayesian additive regression trees (BART) and random forests (RF) to model bird occurrences for 47 species using 43 predictor variables measured at 184 2-ha sites. The majority of predictor variables were in-site habitat variables, but vegetation cover in the surrounding landscape (500 m radius) and geographic coordinates were included to account for known effects of habitat fragmentation and of geographic clines. A consensus model also was developed, built from averaged predictions from the three techniques. We subdivided the avifauna into guilds and other categories (e.g. conservation status) to examine whether there were differences among such subdivisions. Based on cross validation, the consensus model and RF performed best, followed by BART and then BRT. Of the in-site habitat variables, the basal area of red-ironbark trees and groundstorey characteristics such as fineand coarse-litter cover and litter depth had greatest influence on bird occurrences. These results can inform on-site restoration actions (what to restore) and, therefore, complement strategic landscape planning (where and when to restore).

    AB - We sought to identify those in-site habitat characteristics that best predict distributions of woodland birds in the box ironbark region of central Victoria, Australia. Our focus was on comparing and melding outcomes from several forms of ensemble modelling methods, which account for uncertainty in model structure and allow assessments of variable importance. We used boosted regression trees (BRT), Bayesian additive regression trees (BART) and random forests (RF) to model bird occurrences for 47 species using 43 predictor variables measured at 184 2-ha sites. The majority of predictor variables were in-site habitat variables, but vegetation cover in the surrounding landscape (500 m radius) and geographic coordinates were included to account for known effects of habitat fragmentation and of geographic clines. A consensus model also was developed, built from averaged predictions from the three techniques. We subdivided the avifauna into guilds and other categories (e.g. conservation status) to examine whether there were differences among such subdivisions. Based on cross validation, the consensus model and RF performed best, followed by BART and then BRT. Of the in-site habitat variables, the basal area of red-ironbark trees and groundstorey characteristics such as fineand coarse-litter cover and litter depth had greatest influence on bird occurrences. These results can inform on-site restoration actions (what to restore) and, therefore, complement strategic landscape planning (where and when to restore).

    U2 - 10.1111/j.1600-0587.2011.06651.x

    DO - 10.1111/j.1600-0587.2011.06651.x

    M3 - Article

    VL - 34

    SP - 946

    EP - 954

    JO - Ecography

    JF - Ecography

    SN - 0906-7590

    ER -