Discussion
Marine heatwaves are an increasing phenomenon with effects on plankton organisms, food webs, biogeochemistry, and ecosystem services . Thus, the development of reliable forecasting tools on heatwaves properties (e.g., duration, intensity, depth) and ecosystem responses is crucial for successful mitigation and adaptation actions for ocean sustainability . Still, we lack a mechanistic understanding of how temperature relates to the observed ecological alterations during and after the heatwaves. This is mostly due to the strong connection of temperature with various abiotic and biotic factors and the complex temporal and spatial dynamics of marine communities.
In this study, we examine how the temperature anomaly of surface seasonal heatwaves is affecting ecosystem dynamics in plankton communities. We use a trait- and size-structured model that accounts for protists and the life cycle of active and passive feeding copepods. We highlight and discuss three key findings: Firstly, seasonal heatwaves trigger contrasting, lasting effects on plankton communities (biomass, size distribution, functional diversity), extending up to six years post-heatwave onset. Secondly, it is difficult to separate temperature as the key driver of the ecosystem changes we observe when we move from individual processes to community dynamics. Lastly, temperature anomalies can trigger functional groups differentially, with direct and indirect effects varying across groups.
The model results show that the duration of community anomalies depends on when the heatwave occurs. The system takes up to three years for the winter and spring heatwaves and up to six years for the summer and autumn heatwaves to reach the pre-heatwave state. The different mixed layer depth of the seasons (shallow in summer/autumn, deeper in winter/spring) could be a potential driver for this outcome, as it is strongly related to the density of nutrients and plankton in the model and the ocean . Though, since we kept the mixed layer depth fixed in our experiments, we speculate that this outcome is driven by the temperature differences among seasons caused by the heatwaves. The mean seasonal temperature before the heatwaves varies between 15 ˚C and 16 ˚C for winter and spring and 22 ˚C to 24 ˚C for autumn and summer, leading to a temperature difference of up to 9 ˚C between seasons. The seasonal heatwaves of 4 ˚C do not alter the mean annual temperature (20 ˚C). Still, while winter and spring heatwaves maintain temperature fluctuations within pre-heatwave seasonal ranges, summer and autumn heatwaves lead to fluctuations exceeding the pre-heatwave ranges by 2 ˚C (autumn) and 4 ˚C (summer). Thus, the model indicates that the ecological disruption and recovery time are related to the temperature anomaly compared to seasonal temperature fluctuation.
Our results show changes in biomass size bins, dominant groups, and functional diversity index during and after heatwaves, differing across protist and copepod functional groups. Starting with the temperature environmental trait, the model includes functional groups of eight temperature norms. For both protists and copepods, heatwaves lead to changes in the relative biomass and order of the dominant groups during and after the heatwave. Still, functional groups with temperature norms of 20 ˚C and 24 ˚C dominate the plankton community before, during, and after the heatwaves and between seasons.
The community size structure also reflects periodic variations caused by marine heatwaves. However, no consistent pattern emerges across biomass size bins for all heatwave scenarios, revealing the intricate synergy of direct and indirect temperature effects. For years after the heatwaves, the model shows changes in the order of dominant size groups, highlighting that the effect of a seasonal heatwave on the community properties can persist for a long period. However, the reposition of some dominant size groups does not affect the core community size structure. Our model output is supported by previous studies that show that environmental factors beyond temperature likely contribute to size structure variations we observe in nature . In-situ observations have shown that surface heatwaves alter the properties of plankton communities like diversity and size distribution, strongly connected with other environmental conditions such as the passive entrance of species via water masses, stratification, and changes in nutrient concentrations . Field observations show a shorter recovery period than our model projects ranging from a few months to three years depending on the duration of the heatwave . In-vitro and mesocosm experiments also indicate that warming can trigger alterations and different recovery times on physiological rates, species density, and community structure in the same direction as our model.
Our results are in a parallel direction with in-situ andin-vitro observations but are not directly comparative as model-observation disparities stem from differences in design, environment representation, and ecological realism. In-situobservations are snapshots of an ecosystem shaped by many physical, chemical, and biological processes, most of them recorded with a limited temporal and spatial resolution. In comparison to the Eulerian view ofin-situ observations, this study assumes a Lagrangian view and allows us to focus on the theoretical community as moving through time. Mesocosm and laboratory experiments also follow a Lagrangian approach, but they run for shorter periods compared to our model experiments (days to weeks). We also note that descriptive language in most published studies (e.g., small vs big, warm vs cold species) lacks quantitative data (e.g., body size, species temperature optima, and physiological rates) necessary for direct model-observation comparisons. Given plankton’s adaptive plasticity and morphological variations species can manifest as ”cold” or ”warm” depending on regional context, contributing uncertainty to model-observations comparisons.
We propose two more drivers of this model-observational mismatch other than the differences between our model design andin-situ/in-vitro marine heatwaves on the Lagrangian vs Eulerian approach and heatwave properties (e.g., temperature anomaly and heatwave duration): (1) the lack of a 3-dimensional dynamic environment in our model and (2) the need for enhanced ecological and plankton diversity representation. In marine ecosystems, surface heatwaves do not occur in isolation, as in our modelling set-up. They are strongly connected with other abiotic drivers of ecosystem dynamics (e.g., mixed layer depth, nutrient cycling, salinity) that can alter nutrient resources, prey concentrations, and community dynamics during the heatwave . These environmental drivers can also mitigate or aggravate the signal of temperature effect through time. Additionally, even if our model design has complex ecosystem dynamics and higher functional diversity than most ecosystem models (Petrick et al., 2022), it does not consider phenotypic plasticity, evolution, or behavioral decisions like vertical migration and changes in foraging and predation avoidance that can maximize fitness on an individual level and resilience on a community level . These physiological and behavioral responses might allow species persistence that could dampen the impact of the heatwave and accelerate recovery periods.
Our study highlights the essential role of functional diversity in population dynamics. In the model, protists are the most diverse community with plasticity on energy uptake and short life cycles. They show dynamic responses to environmental conditions and experience more changes in community composition compared to copepods. Passive copepod feeders are more vulnerable than active feeders and have the longest recovery time in terms of biomass concentration. This is probably due to their trade-off disadvantage on resource competition combined with predation losses from active feeders. Traditionally, research has focused on prey density and properties, but studies have shown that the modes of energy uptake and foraging also have a strong impact on ecosystem dynamics, biogeography patterns, and biogeochemistry . For example, studies have shown that mixotrophy evolved as a survival strategy against prolonged periods of darkness and that some copepod species can actively switch between passive and active feeding depending on the environmental conditions (Kiørboe et al., 2018a). Closer attention to the feeding mode and organismal behavioral decisions that lead to trait optimization and fitness can help us to better understand the community status in different environmental conditions and extreme events. A gradual increase of functional diversity in the model could provide us with a new level of mechanistic understanding of ecosystem dynamics but probably increase the difficulty of distinguishing drivers. It could also highlight suggestions on data needs for advancing the state-of-the-art of forecasting tools and our confidence in trustworthy projections crucial for policy advice and actions. Our model is a useful tool for mechanistically exploring the effect of abiotic parameters in ecosystems from individuals to community levels. We wish to see future studies using and adjusting the model design to explore the resilience of plankton communities under different environmental conditions.