Managing ecosystems in an increasingly variable world

Global climate change is leading to a more variable and unpredictable environment in many places. The project ECOVAR investigates how this important aspect of climate change will affect individuals, populations and ecosystems. Previous studies have indicated that both positive and negative responses to variability are possible, but the underlying mechanisms involved are still poorly understood for most species. The main goal of ECOVAR is to advance our understanding of effects of variability through a combination of theoretical modeling, comparative analyses, and a detailed experimental study. The laboratory experiment was conducted with the water flea species Daphnia magna. Water fleas play a key role in freshwater ecosystems, as filter feeders and as food source for other organisms. The experiment was done with individuals from four clonal lines, at eight treatments; six constant temperatures, and two variable temperatures (daily fluctuations around a cold and warm mean). Fundamental life history responses were recorded for each individual: growth rate, age at maturity, moulting rate (development), survival, and reproduction (number and size of offspring per clutch). Thermal variability affected all responses, with overall positive effects at cold temperatures, and negative at warm temperatures. These effects were also more positive overall than predicted from non-linear averaging of responses at constant temperatures, reflecting that this organism is well adapted to variable conditions. The data on individual performance will be incorporated in structured population models, which will be used to compare the mean fitness in the variable and constant environment, and identify the key demographic parameters mediating the difference. The project also contributed theoretical development to improve our understanding of the long-term persistence of species in variable environments, and different evolutionary adaptations depending on life history. Previous studies are largely based on models without an explicit link between demographic parameters and climatic drivers (e.g. temperature), implicitly assuming linear relationships which leads to predicted effects of variance on persistence always being negative. However, non-linear relationships can alter this conclusion, and positive responses may be possible when some vital parameters show a convex relationship with the climate variable. A key challenge with structured populations (where individuals have different age or stage) is to assess the overall effect of nonlinearity in different demographic rates across the life history. We developed a new index for the total curvature in a life history, which predicts an important part of species responses to variability. This theoretical result provides the foundation for a large comparative analysis which constitutes another main part of the project. Here, the goal is to identify how different adaptations to variable environment vary across the fast-slow continuum of life histories. Using simulations and analyses of stochastic structured models developed for animal species in the database COMADRE we investigated responses of the long-term population growth to climate variability. We used multiple filtering criteria to ensure high quality comparisons and selected 81 populations from the database (2 amphibians, 20 birds, 2 bony fish, 39 mammals and 18 reptiles). The scenarios considered how age-specific parameters are affected by the focal climate driver (types and degree of curvature), and different kinds of covariance between survival and reproduction. The results show that short-lived species show the strongest potential for population growth responses to variability, whether the effect is positive or negative. The direction of the effect depends on the total curvature of vital rate responses and their covariance; positive effects of variability are mostly found for short-lived species with a positive total curvature, and with negative covariance between survival and reproduction. The ECOVAR project has contributed new methods required to study effects of climate change in structured populations. One study highlights the use of hazard rates to model effects of environmental drivers on e.g. survival. In another, we provide methods for sensitivity analyses of stochastic matrix models, relevant to study effects of climate variability. Demographic models from the project have also been applied to study fitness consequences of climate change and variation in other study systems, including cod, pike, and springtails. Better knowledge of these mechanisms is also essential to improve conservation and management strategies.

Prosjektet bidrar med ny kunnskap om hvordan arter responderer på slik variabilitet, og hvordan responsene avhenger av artenes livshistorie, f.eks. om de er kortlevde eller langlevde. Disse resultatene vil kunne bidra til bedre strategier for forvaltning og bevaring av arter i områder som opplever endringer i variabilitet og forutsigbarhet av miljøet, en betydelig konsekvens av klimaendringene. Metoder og modeller som er utviklet i prosjektet er allerede anvendt i andre studiesystem, og har potensial for å kunne brukes av flere forskere og på flere studiesystemer. Prosjektet har videre bidratt til karriereutvikling for de sentrale prosjektdeltakerne, samt økt internasjonalt samarbeid. Flere masterstudenter har også gjennomført sine prosjekter på temaer tilknyttet prosjektet og med veiledere fra prosjektet.

Global climate change is altering not only long-term average conditions, but also the short-term variability of environmental drivers such as temperature and precipitation. We know that climate change will affect populations and ecosystems, but ecological research has so far mainly focused on changes in means. Knowledge of the effects of changing variability is much more limited, including potential interactions with changes in the mean. Understanding how environmental variability affects rates of population growth is of fundamental importance to ecological and evolutionary processes like changes in abundance, trait distributions, species composition and interactions, and is also essential to develop management strategies advocated to deal with climate change. Theory predicts that both positive and negative effects of increased variability on long-term population growth are possible, but the underlying mechanisms to such responses are still poorly understood for most species. This project will advance our understanding of biological responses to climate variability, through a synergistic combination of approaches of theoretical modeling, comparative empirical analyses, and a detailed experimental study. In order to identify key ecological, geographical, and life history characteristics that can predict population responses to changes in climate variability, we will analyse a database of demographic population models on plants and animals (COMPADRE / COMADRE). The experimental study will provide detailed insights into underlying vital rate responses to changing variability, for a case where a mean/variance interaction is expected. Finally, to assess ecosystem consequences of changing variability together with other stressors such as harvesting, we will develop and analyze stochastic food web models, informed by the results of the project. Knowledge generated by this project may alter current conclusions as well as management strategies.