INTRODUCTION
One of the most widespread consequences of climate change is the global redistribution of life on Earth (Chen et al. 2011; Poloczanskaet al. 2013; Pecl et al. 2017). The climate-mediated movement of species into novel environments alters species assemblages and this can have flow-on effects for ecosystem functioning and services (Pecl et al. 2017). Ecological impacts of species redistributions can be particularly severe when foundation species are affected, as this can have cascading impacts on associated species and lead to regime shifts, whereby one ecosystem state shifts to an alternative state (Hughes et al. 2013).
Climate-driven species redistribution models generally predict net biodiversity declines and biotic homogenisation at a global scale (Magurran et al. 2015; Savage & Vellend 2015). However, following changes in environmental conditions, species diversity can initially increase at the local scale (Hiddink & Ter Hofstede 2008; Walther et al. 2009). Emerging evidence shows that patterns of increased diversity are often masked by biological processes not easily observed, including lag times and transitional stages (Daskalovaet al. 2020) and/or observations reflecting a turnover of species where numbers of “winners” and “losers” are similar and diversity appears to be maintained where lost species are offset by new ones (Dornelas et al. 2019; Finderup Nielsen et al. 2019). The need for long-term datasets to uncover assemblage changes is therefore paramount for accurately predicting ecosystem change.
There are large-scale coordinated efforts to document the arrival of range-expanders into new areas (Robinson et al. 2015), but we know relatively little about their interactions with and responses of recipient communities (Bates et al. 2014; Svenning et al.2014). Additionally, we lack understanding of how climate-mediated losses of species, and the traits associated with those species, affect ecosystem processes (Wardle et al. 2011). Such understanding is critical to the development of management tools for mitigating impacts of range expanding organisms on ecosystem function and services.
Biogeographic transition zones where abiotic and biotic conditions allow for the interaction of organisms from different geographical origins are experiencing rapid changes in the marine realm (Horta e Costa et al. 2014; Troast et al. 2020). Transition zones provide a unique opportunity to understand the impact of range shifting organisms because populations of species from the surrounding environmental extremes already exist and interact. Transition zones allow us to examine the processes and mechanisms that facilitate successful species expansions, as well as those which make resident species resilient within their native range despite the establishment of range-expanders. The identification of mechanisms that facilitate changes in abundance for some species under these conditions may help identify range expanders and/or at-risk species from the surrounding systems and for similar systems elsewhere. For example, recent evidence from tropical/temperate transition zones has shown that, despite historical mixing of tropical and temperate species, the increasing abundance of range-expanding tropical herbivorous fishes can overgraze seaweeds and lead to regime shifts by maintaining reefs in canopy-free states (Vergés et al.2014; Bennett et al. 2015).
In recipient systems where range-expanding species are becoming more common, predicting coexistence with resident species is critical to understand the eventual structure and functioning of those systems. Mechanisms that underpin coexistence are complex and influenced by multiple ecological and evolutionary processes acting simultaneously (Pinsky 2019). However, the well-established field of invasion ecology can be used as a starting point to make predictions about climate-driven species redistributions (Wallingford et al. 2020) whereby the trait space of range expanders can then be empirically quantified (Sunday et al. 2015). For instance, trait-based analyses of invading species are often used to predict the likelihood of competitive interactions with resident/native species and effects on the recipient community (van Kleunen et al. 2010). Invasive species often show generalist traits that facilitate success in novel environments, often to the detriment of resident specialist species (Clavel et al.2011). Dynamic environments often favour generalist species that more readily adapt (Kassen 2002) and so we could expect that, in a range expansion context, resident specialist species may be the most vulnerable.
For transition zones affected by climate change, species coexistence may continue to be possible if incoming species show unique traits (Paciogluet al. 2020), i.e. do not directly compete with residents or are able to show plasticity in resource use in a novel setting (Jackson & Britton 2014). However, there are cases where expression of unique traits may not be beneficial if the system is already disturbed/degraded (Bulleri et al. 2020). Alternatively, resident populations may be far below carrying capacity (for example due to fishing in marine systems), and thus even if incoming species substantially overlap in trait (niche) space, competitive exclusion will not occur and coexistence between migrants and residents can persist.
On the east coast of Australia, the Solitary Islands region is a sub-tropical transition zone within a climate change hotspot (Hobday & Pecl 2014). This region is strongly influenced by the East Australian Current, a western boundary current that is strengthening in response to climate change and which can facilitate the poleward range expansion of marine species (Ridgway & Hill 2012; Castro et al. 2020). An increasing abundance of tropical herbivorous fishes has been linked to the decline of temperate habitat-forming kelp in this region (Vergés et al. 2016). The kelp Ecklonia radiata is a critical component of Australia’s temperate reefs that supports ecologically and economically important fish and invertebrate species (Wernberg et al. 2019). The loss of seaweeds in eastern Australia is part of a global “flattening of kelp forests” (Filbee-Dexter & Wernberg 2018) whereby canopy seaweeds are replaced by low biomass turfing algae or corals (Vergés et al. 2019). How this shift in benthic cover affects resident fish communities in a tropicalisation context is largely unknown, however declines in kelp-dependent species may be expected concurrent with the decline of this foundation species.
Here, we surveyed tropicalised reefs within the Solitary Islands region to quantify changes in fish assemblages during a period of kelp loss. In particular, the study period encompasses a shift in benthic cover where temperate kelp Ecklonia radiata was completely lost, did not recover, and was largely replaced by turfing algae and corals (Vergés et al. 2016). We analysed baited remote underwater video surveys (BRUVS) collected over 17 years from 2002-2018, with total kelp loss occurring from 2009 (Vergés et al. 2016). Abundance and morphological traits of individual fish species were assessed. We used the morphology of individual fish species within the community to investigate the potential for trait differentiation that may be indicative of meaningful competitive interactions sensu (Azzurro et al. 2014) between tropical and temperate associated species of similar feeding guild (herbivores, invertivores, piscivores and planktivores). With particular interest in the vulnerability of temperate species, we specifically asked: 1) How has the probability of occurrence of tropical and temperate fishes from different trophic guilds changed over 17 years, encompassing the transitional period from kelp to turf dominated reefs? 2) Are there any species, tropical or temperate, that have significantly increased and/or decreased in abundance over this period? Based on the analysis of fish morphological traits, 3) is the overall “trait space” of incoming tropical fishes different from that of resident temperate ones, and 4) are changes in fish abundance linked to unique morphological traits?