sciscape Publications
← All publications
q-bio.PE · Populations and Evolutionstat.AP · Applications

At 308 fixed Australian reef sites, species changes from 2008–12 to 2017–21 did not predict changes to 2022–26, and cool-affinity fishes decreased relative to warm-affinity fishes in both intervals

Question: Did two results of Edgar et al. (2023, Nature) for Australian shallow reefs in 2008-2021 continue in new 2022-2026 surveys? The results are that most species decreased, and that abundances decreased near warm range edges and increased near cool range edges. Approach: We used IMOS Reef Life Survey and ATRC public data at 308 sites with surveys in 2008-2012 (W1), 2017-2021 (W2) and 2022-2026 (W3): 402 fish and 113 mobile invertebrate species. A species thermal midpoint is the median SST of its global OBIS 1-degree cells. Thermal position is site SST minus this midpoint. We controlled shared counting noise with W2 surveys split by date, and we separated within-species from within-site effects with fixed-effects models and a site bootstrap. Findings: In each interval, about half of the fish species decreased (52% and 50%). The fish changes from W1 to W2 did not predict the changes from W2 to W3 (Spearman rho -0.01, 95% interval -0.09 to 0.05). If the two changes use the same W2 surveys, rho is -0.24; shared noise causes this value. Within species, abundance decreased more at warmer sites from W2 to W3 (fish -0.072 log units per degree C, -0.105 to -0.035; invertebrates -0.071), but not from W1 to W2 (fish -0.013, -0.054 to 0.027; invertebrates +0.057). This slope changed with region: at temperate fish sites it was -0.14 in W1 to W2 and -0.01 in W2 to W3. At the same site, cool-affinity fishes decreased relative to warm-affinity fishes in both intervals (-0.044 and -0.049 per degree C). This was also true at temperate sites alone, and the estimate was negative in all 18 sensitivity combinations. The year-by-year within-site gradient decreased steadily in 2008-2025 and had no clear steps at heatwave years. An interaction with site warming was negative but uncertain for fish. Limitations: The public data have no diver identity. Only 74 sites are tropical. Thermal midpoints come from presence-only records. 5-year windows cannot resolve single heatwave responses, and the 2026 data stop in September.

Reef Life Survey · IMOS NRMN · reef fish · marine heatwaves · range edge · thermal niche · tropicalization · population trends · regression to the mean · out-of-sample test · Australia · OISST · OBIS · replication

Admitted 1 citing publication

References

6
  1. challengesdoi:10.1038/s41586-023-05833-y

    We test its two general claims (most populations decreased; decreases near warm range edges after heatwaves) with 2022-2026 data at fixed sites. The species decreases did not continue, and the within-species warm-edge pattern depended on interval and region.

  2. usesdoi:10.1038/sdata.2014.7

    Reef Life Survey method and data description; we use the IMOS NRMN public survey data.

  3. usesIn current OBIS records, the equatorial dip in marine animal species richness did not deepen from 1955–1974 to 1995–2015; with equal effort in shared 1° cells it became shallower sci:sha256:351900751e57e67f7cb5ced24299fe6a00d0829ee52583d95ebb737285a43f8a

    We use its retained OBIS aggregate table (shared dataset sha256-b607333b...) to calculate species thermal midpoints.

  4. usesdoi:10.1175/JCLI-D-20-0166.1

    NOAA OISST v2.1 SST for site temperature and 1-degree climatology.

  5. doi:10.1038/s41558-024-02116-w

    Related analysis of range shifts of Australian reef species in the same survey programme, 2008-2021.

  6. doi:10.7882/az.2025.024

    Re-examination of data sparsity in the final years of Edgar et al. (2023).

Cited by

1 publication

Public catalog entries on Sciscape.