Research

Ph. D. research: Adaptations to variable and unpredictable environments

Anthropogenic climate change is causing rapid changes in living conditions for organisms around the globe. Importantly, the environments organisms experience are changing not only in their mean conditions (e.g. becoming warmer, wetter, drier...), but also in how variable or predictable conditions are. There has been a lot of work done on how organisms should deal with such variation on different time scales. While the field of behavioral ecology has typically focused on short-term variation and within-generation adaptations (e.g. insurance, risk-sensitivity), this has seldom taken a long-term view and considered that conditions may change between generations as well, affecting the optimal behaviors in an individual's lifetime. Such problems have rather been dealt with by evolutionary biologists, who have identified key concepts such as bet-hedging strategies and phenotypic canalization. However, these various adaptations to short- and long-term environmental variability have rarely been placed in the same framework, and it is not always clear how they relate to each other, whether and how they will interact, and to what extent they are even mutually compatible.

The approach of my Ph.D. project is therefore to use a variety of theoretical modelling tools to link short- and long-term adaptations to uncertainty and unpredictability. I will take well-established concepts within behavioral ecology, and apply an evolutionary perspective where the environment varies both within and between generations. Below is a summary of some of my current and past projects.

Insurance, bet-hedging and canalization: Phenotypic evolution in a variable environment

In the first paper of my Ph.D., I use skewed fitness functions to investigate the effects of insurance and conservative bet-hedging in coping with within- and between-generation environmental variability. I show that the relative importance of conservative bet-hedging depends on the amount of diversifying bet-hedging (adaptive phenotypic variation among offspring), and confirm and expand on previous results from the theoretical literature. This work is currently available as a preprint on biorXiv, and is currently in revision for Evolution (Haaland, T.R., Wright, J., Tufto, J., Ratikainen, I.I. 2018).

With: Irja Ida Ratikainen, Jarle Tufto, Jonathan Wright.

Acknowledgements: Luis-Miguel Chevin and Andrew Simons for valuable discussions.

I've been told my fitness landscapes look like Edvard Munch's The Scream.

Specialist, generalists and bet-hedgers: Phenotypic variability within and between individuals

Using the same framework as in the above paper, I investigate how within-individual phenotypic variability (which can be seen as for example niche width) affects the adaptive advantage of between-individual variability, such as diversifying bet-hedging. This work is still in preparation, but promises to reveal exciting insights. I identify different strategies depending on whether the traits we are interested in cause fitness to accumulate additively (for example in foraging related traits) or multiplicatively (for traits directly affecting survival probability), and depending on how the environmental conditions vary within and between generations. In what cases might generalism represent a bet-hedging strategy? When is it better to produce very different (bet-hedged) specialist offspring, than very similar generalist offspring?

With: Irja Ida Ratikainen and Jonathan Wright

Acknowledgements: Lovely, underappreciated papers by Michael Lynch, Wilfred Gabriel and George Gilchrist.

Snippet from my poster presentation given at Evolution 2018.

Adaptation to rare events: What can evolutionary history tell us about vulnerability to climate change?

Instigated during a research stay at Dr. Carlos Botero's lab at Washington University in St. Louis, USA, I have built a series of models on the relative importance of additive and multiplicative fitness accumulation in evolutionary processes. Additive fitness accumulation occurs in traits that are subject to many selective events in each individual's lifetime, and in heterogeneous environments where individuals in a populations experience different conditions in the same generation. Conversely, multiplicative fitness accumulation occurs in traits that are subject to one or few selective events in each lifetime, and in homogeneous environments where all or most individuals in a population experiences the same conditions each generation. These different modes of fitness accumulation can select for very different traits in the long term, and we are interested in how these different traits can leave populations differentially vulnerable to ongoing global changes in extreme weather events such as floods, droughts, wildfires and hurricanes.

During my stay in Dr. Boteros' lab I also helped develop experimental procedures to test these predictions empirically, on the yeast Saccharomyces cervisiae. These experiments are currently under way. Combined with results from our models where climate change leads to different types of changes in the regimes of extreme events, we hope that this work can reveal new insights for setting conservation priorities in the face of anthropogenic climate change.

With: Suchith DaSilva, Justin Fay, Carlos Botero. 

Acknowledgements: Vince Fasanello and Ping Liu for work in the yeast lab; Jeremy Van Cleve for theoretical input.

Pie colors depict proportion of different evolutionary outcomes at given parameter combinations.
Some scenarios are very hard to adapt to (small or unexisting pie charts).

How do you track a changing environment over time, and when should you not bother? 

I have worked on simulation models exploring the evolution of learning and sampling in a variable and unpredictable environment. Climate change is causing problems for organisms that rely on environmental cues such as photoperiod or temperature. Cues that previously were reliable indicators of when to start breeding are now causing phenological mismatch for species in many parts of the world. Animals can also base their decision on own experience - what worked well last year? If conditions are likely to be similar to the last time the decision was made (temporal autocorrelation), relying on memory may be a good strategy.

Here I allow genes for plasticity, sampling effort and memory to co-evolve under different degrees of temporal autocorrelation and cue reliability. I am then interested in exploring how well the evolved strategies perform when one of these environmental parameters change - which is what humans are doing to animals now. By making weather more extreme and unpredictable, by warming the planet and shifting the onset of spring, by removing or decoupling links between cues used by organisms and fitness-relevant information, we are posing novel challenges to animals around the world, and it is therefore important to know how or whether they will be able to keep up with these new conditions in the future.

With: Jonathan Wright and Irja Ida Ratikainen.

Acknowledgements: Carlos Botero, Trevor Fristoe and Aimee Dunlap for helpful discussions.

Game theory, behavioral reaction norms and information use in a variable world

Game theory concerns itself with situations where two or more individuals (players) interact, and where the best decision for each player depends on the actions of the other player(s). While much has been said and done about this rich and fascinating field - also within the realms of animal behaviour - a predictive theory about how variable environments affect individual decisions has been lacking. I am interested in exploring the evolution of social behavioural reaction norms, that is, how an individual should behave depending on its social environment.

So far I have been focusing on a social foraging case known as the Producer-Scrounger game, but the concepts are easily extended to other game theoretical problems. I go beyond traditional ESS theory in which stable frequencies of strategies within a population are well established, to investigate the strategies (reaction norms) of the individuals that make up the population. When the population is at the ESS, all individuals by definition do equally well. But how does the population arrive here? Do certain individuals always use one or the other strategy, regardless of what the others are doing? Or are they plastic, i.e. always responding to their social environment? If so, how do they gather this? Is private information better than public information in deciding which tactic to use? These are only some of the questions that can be addressed with this novel approach to a well-known field of research.

With: Jonathan Wright, Irja Ida Ratikainen, 

Acknowledgements: Håkon Johansen for getting the project a flying start with his modeling and writing; Arnon Lotem, Aimee Dunlap and Carlos Botero for valuable discussions.

What's the optimal shape of this reaction norm?
And does it depend on environmental factors, information use or learning mechanisms?

Other projects

Sexual selection and parental care 

For my Master's degree at NTNU I created a state-based stochastic dynamic model of the differential allocation hypothesis - that parents adjust their investment in offspring depending on the quality or attractiveness of their current mate. I have continued working on this topic on the side of my Ph. D. work, as it is an intriguing problem attracting a lot of attention from behavioural ecologists and evolutionary biologists alike.

This work has resulted in two publications, one presenting my stochastic dynamic model alongside an analytical model which supports our results (Haaland, T.R., Wright, J., Kuijper, B., Ratikainen, I. I.; 2017 Am. Nat., open access!), and one where we extend the stochastic dynamic model to investigate how mate quality can affect the trade-off between offspring size and number (Ratikainen, I. I., Haaland, T. R., Wright, J.; 2018 Proc R. Soc. B., also open access).

I have co-supervised an MSc project testing the logic behind Burley's original formulation of the differential allocation hypothesis: That investing more when you're with a high-quality mate is beneficial because it increases the chance of this mate staying with you until the next breeding season. While this argument has all but been abandoned in discussions of differential allocation since, we tested whether the predictions from our Haaland et al 2017 model still hold when allowing for this effect. In order to critically test the conclusions from this model, we are currently looking into an adaptive dynamics model of the evolutionary feedbacks between mate choice and differential allocation at the population level, in collaboration with Eva Kisdi and Tadeas Priklopil.

Acknowledgements: Hanna Kokko, Andy Higginson, Franjo Weissing.

Male pipefish Syngnathus typhle brood the eggs that females deposit in their mouth.
Turns out, females lay larger eggs if paired with a smaller male. We explain why.

Arctic fox (Vulpes lagopus) behaviour and conservation

In collaboration with the researchers from Stockholm University in Sweden, I have been gathering field data on the critically endangered Scandinavian Arctic foxes and assisted with conservation measures such as tagging, monitoring and sampling foxes and cubs for population inventory, vaccination, supplementary feeding, assessing habitat quality (vegetation analysis) and food availability (bird inventory, rodent trapping to monitor long-term cycles). Observing the foxes day and night for months during the Arctic summer, I am now interested in developing models to examine the reproductive decisions, territorial behavior, division of labour, dispersal and predation risk of Arctic fox families. Hunted to near extinction about 100 years ago, Arctic foxes have failed to recover after their protection, as they are faced with a number of threats, such as inbreeding depression due to low population sizes, low connectivity between subpopulations, diseases and competition from red foxes and lower food availability due to climate change. Arctic fox families are facing different challenges than before, and understanding their behavioural dynamics can therefore help us predict how they will react to future conditions, and allow us to make more effective conservation measures.

With: Anders Angerbjörn, Rasmus Erlandsson, Zoologiska Institutionen, SU.

Summer under the midnight sun spent together with these guys
is nothing less than inspiring, soul-enriching and wonderful.

Using IT tools for teaching in field courses

I have been teaching botany labs and field trips on the undergraduate course Floristics and Faunistics in Norwegian Ecosystems in 2015-2017, and have designed an experiment to test the short- and long-term learning outcome of using Geographic Information Systems (GIS) in the field.

With: Kristiane Midtaune, Jakob Bonnevie Cyvin, Jan Ketil Rød, Department of Geography, NTNU; Sigrid Lindmo, Ragnhild Thorsen Grevskott, Department of Biology, NTNU.