Climate-driven invasion and incipient warnings of kelp ecosystem collapse

Climate change is causing widespread species redistribution and ecological change, leading to predictions of nonlinear, catastrophic ecosystem collapse with severe consequences. The ability to detect early-warning signals is crucial to mitigate these impacts. Traditional early-warning signal detection relies on high-frequency, long-term time-series data, which are often unavailable for most ecosystems. This study proposes that spatial pattern formation, as warm-environment species expand their range poleward, may serve as an early warning. Southeastern Australia, experiencing accelerated ocean warming due to the East Australian Current (EAC), serves as a case study to investigate this hypothesis. The strengthened EAC has driven poleward range extensions of approximately 70 marine species, notably the long-spined sea urchin (*Centrostephanus rodgersii*), a key driver of kelp forest decline through overgrazing. This introduction sets the stage for examining the spatial patterns of kelp forest collapse and the utility of ‘incipient barrens’ as early-warning signals.

The concept of 'critical slowing down' in high-frequency ecological time-series data has been central to early-warning signal detection. However, the required data frequency and length are often unattainable. Previous research has highlighted the impact of climate change on species redistribution and subsequent ecological cascades (Pecl et al., 2017; Vergés et al., 2014). The potential for catastrophic shifts and alternative stable states in ecosystems has been extensively studied (Scheffer et al., 2001, 2009, 2012; Rietkerk et al., 2004), with emphasis on identifying early warning signals before irreversible change (Hughes et al., 2012). The study builds upon prior work documenting the range extension of *C. rodgersii* in Tasmania (Ling et al., 2008, 2009; Johnson et al., 2011; Gervais et al., 2021) and its impact on kelp forest ecosystems (Ling et al., 2009, 2012, 2015; Johnson et al., 2013). The literature review establishes the context for using spatial pattern formation as an alternative approach to detecting early warnings of climate-driven ecosystem collapse.

The study utilized a space-for-time substitution approach, leveraging a 15-year period (2001/02 and 2016/17) of data from 13 east coast Tasmanian reef sites. Data were collected using spatially hierarchical surveys: 1) Diver surveys to 18 m depth assessed urchin abundance (*C. rodgersii* and *Heliocidaris erythrogramma*), lobster and abalone densities, and substratum types (156 transects). Barrens cover was estimated within 5 m² quadrats. 2) Towed-underwater-video surveys (80 km per period) covered 4-40 m depth to assess broader patterns of barrens cover, categorized into four types: continuous and three sizes of incipient barrens. Data analysis involved ANOVA with nested effects of time, site, and subsite, assessing the changes in urchin abundance and barren coverage over time. Multiple regression was used to examine the relationship between urchin density/barrens cover and various abiotic (reef type/depth) and biotic (predatory lobsters, native urchins, abalone) factors. The space-for-time substitution is achieved by comparing the earlier and later surveys, treating the spatial distribution of the range extension as a temporal progression of ecosystem collapse. This detailed methodology section explains the comprehensive data collection and analysis techniques used in the study, emphasizing the spatial and temporal scales considered and justifying the space-for-time substitution.

The study revealed a significant increase in *C. rodgersii* abundance (x1.8 overall, x1.7 in eastern Tasmania) and barren cover (x3.9 in eastern Tasmania) over the 15-year period. Urchin abundance increased disproportionately more slowly than barren cover, indicating a steepening negative trend between kelp cover and urchin density over time. This suggests kelp forests are increasingly vulnerable to collapse once a certain urchin density threshold is exceeded (overgrazing tipping-point). Barren increase was greatest on deeper reefs and large boulder reefs, where urchins find refuge from predators and wave action. Multiple regression showed a positive association between *C. rodgersii* and native urchins, and negative associations with predatory lobsters and abalone. The analysis of incipient barrens revealed a dramatic increase in their prevalence over time, acting as an effective early-warning signal of impending collapse. Projections based on the observed expansion rate indicate that ~50% of eastern Tasmanian reefs could collapse to barrens by 2029. The findings highlight the utility of incipient barrens as an early-warning indicator of catastrophic phase shifts and provide quantitative evidence of the spatial and temporal dynamics of kelp forest collapse in a warming ocean.

The findings support the hypothesis that spatial pattern formations, specifically the emergence and coalescence of incipient barrens, serve as effective early warning signals of climate-driven ecosystem collapse. The space-for-time substitution approach allowed the detection of these early warnings well in advance of widespread collapse. The strong negative association between urchin abundance and the presence of predatory lobsters and abalone highlights the role of trophic cascades and overfishing in reducing ecosystem resilience to climate change. The observed uneven distribution of urchins and barrens across sites and depths underscores the complex interplay of abiotic and biotic factors influencing the invasion and overgrazing processes. The study's findings have direct implications for management strategies, emphasizing the need for proactive climate adaptation measures to protect kelp forest ecosystems. The results provide a powerful demonstration of the value of spatially extensive monitoring for detecting early warnings of climate-driven regime shifts.

This study demonstrates the effectiveness of using space-for-time substitutions to detect early-warning signals of climate-driven ecosystem collapse. The emergence and expansion of ‘incipient barrens’ provide a clear indication of impending catastrophic phase shifts in kelp forests. The results highlight the critical role of trophic cascades in shaping ecosystem resilience and emphasize the need for proactive management strategies, such as rebuilding predator populations and implementing targeted urchin harvesting, to mitigate the impacts of climate change. Future research could focus on refining the predictive power of the incipient barrens indicator and exploring the effectiveness of different mitigation strategies in various ecological contexts.

The study focuses on a specific geographic region and species, limiting the generalizability of the findings to other ecosystems. The space-for-time substitution approach assumes a consistent rate of change over time, which might not always hold true. The study's reliance on visual surveys might introduce some observer bias, although efforts were taken to minimize this using standardized methods and multiple observers. Further research across different ecosystems and species is needed to validate the applicability of the incipient barrens as a universal early warning signal. Finally, although mitigation efforts are mentioned, the long-term effectiveness of these is still to be confirmed.