How well do we understand the way body size affects populations?

We have all heard the saying “live fast, die young” but it doesn’t only apply to film stars; smaller life forms also abide by this rule. Microscopic phytoplankton cells can double in numbers every few days, while the much larger elephant lives almost 100 years and reproduces slowly. This relationship between body mass and the ‘pace of life’ is well known. But the underlying mechanisms are far from resolved.

A number of years ago, scientists proposed a theory to link the ‘pace of life’ of individuals to the ecology of populations, communities, and ecosystems. The Metabolic Theory of Ecology predicts how energy use (metabolism) is affected by body size and how growth rate, development time and death rates change with energy use. These changes in individual rates will, in turn, affect rates of change in populations.

Metabolic Theory is very appealing because it explains many patterns we see in nature and is based on a fundamental rate which applies to all levels of organisation. But it is difficult to test. Detecting differences in energy use accurately requires having consistent differences in body sizes among organisms, while controlling for other variables (e.g. age, nutrition, health). This usually means comparing across species, which differ in more ways than just size.

The artificially evolved phytoplankton were almost 400 generations old when this work was done. The graph shows the considerable differences in size between small and large algae which allowed Martino and Dustin to ask their questions about body size, metabolism and populations within a species.

Few tests of these predictions have been able to control for these variables but recently Martino Malerba and Dustin Marshall were able to do this and they found that the body size-metabolism relationship did not predict population change as expected.

Martino and Dustin were in the enviable position of having access to artificially-evolved large and small phytoplankton cells of the same species that differed in size enough for them to ask “what happens to the population of a species if the average body size of individuals change?” They could see how body size affected energy use, growth rates, density and biomass.

It turns out that the story was far more complex than expected. The effect of body size on how organisms use energy and grow was very strong but also varied during the course of evolution.

So why didn’t body size explain trends in growth and energy use among size-evolved organisms? The answer may lie with previous work from the Centre for Geometric Biology. Martino and his colleagues found that large-evolved plankton cells optimized their photosynthetic pigments and produced more energy overall than smaller cells. This suggests larger cells have greater access to resources than smaller cells and so the way in which body size and metabolic rate influence the demography of a species is not as predictable as we once thought.

Instead of the classical view, where body size determines the rate at which organisms use energy, which then determines demographics, Martino and Dustin suggest that body size can affect metabolism and populations at the same time.

IIf these laboratory cultures are representative of natural populations, we would predict that current trends of reduced body size (from global warming) could lead to lower rates of population increase, biomass productivity and maximum biomass. This is the opposite of what current theory would predict. This is particularly important when we consider the role of phytoplankton in fixing carbon and supporting food chains.

This research was published in Ecology Letters.

How does size, fragmentation and food affect metabolic rates in a bryozoan?

We know that metabolic rate (a measure of energy use) tends to vary among individuals of different sizes and also with food availability. Lukas Schuster has been working with his PhD supervisors Craig White and Dustin Marshall to find out how metabolic rate changes with size in the colonial marine invertebrate Bugula neritina. They are also interested in how (or if) manipulating an organism’s size affects metabolic rate in individuals that were either starved or fed.

Lukas and his supervisors chose to work with Bugula because it is a colonial organism. These types of animals can be used for testing metabolic theories because the size of the colonies can be changed by removing fragments.

Lukas measured metabolic rate by measuring oxygen consumption of intact and size-manipulated colonies of Bugula that had been fed or starved as well as colonies that he had grown up in the field so he knew how old they were.

When Lukas measured metabolic rates for individuals of known age he found that rates increased proportionally with size across the different ages, that is, metabolic rate scaled isometrically with size when looking at individuals that ranged in age / developmental stage.  In contrast, when he looked at metabolic rates for a specific age or developmental stage he found that rates didn’t scale proportionally with size but, instead, had allometric scaling as has been found in previous studies.

Lukas and his supervisors point out that it is important to be aware of these differences. Measuring metabolic rates of field-collected specimens of unknown age may result in isometric scaling of metabolic rate with size. Conversely, measurements of specimens at the same developmental stage is likely to result in allometric scaling where larger individuals have proportionally lower metabolic rates compared to smaller individuals.

To the team’s surprise, they also found that when they measured metabolic rate in size-manipulated Bugula that has been collected from the field, metabolic rate reverted to allometric scaling. Manipulating size in Bugula may lead to a leaking of nutrients through the pores between the zooids that make up the colony and this may be driving the change in the relationship of metabolic rates with size.

Bugula responded to food deprivation by reducing its metabolic rate, and conversely responded to feeding by increasing its metabolic rate, which was consistent with what other researchers have found in other species. But, in comparison to other species, the rate at which Bugula increased its metabolic rate following feeding, was rather low. This may also relate to the fact that Bugula is a colonial species but as there are very few studies investigating the way metabolic rate responds to feeding in colonial organisms, it is hard to know for sure.

Clearly, the relationship between size and metabolic rates in Bugula is complicated and may relate, in part, to the fact that Bugula is a colonial organism. But to fully understand the effects of size manipulation on metabolic rates and biological processes within Bugula colonies, further studies will be needed.

This research is published in the journal Invertebrate Biology.

Why do cooler mothers produce larger offspring?

In a recently published letter, Amanda Pettersen, Craig White, Rob Bryson-Richardson and Dustin Marshall propose a simple model to explain a pervasive conundrum – why do cooler mothers produce larger offspring?

Life history theory maintains that mothers balance the costs and benefits of making a few larger and better performing offspring against making many smaller and poorer performing offspring.

A major challenge to the theory is the fact that temperature seems to alter the optimisation of this trade off. Observations indicate that across a wide range of taxa and systems, mothers in warmer conditions produce smaller offspring. What is more, experimental studies have also shown that increasing temperatures decrease offspring size.

Amanda and her PhD supervisors are proposing that linking life history theory and metabolic theory, which relates to energy use, can provide a widely applicable explanation to the offspring size / temperature relationship.

Their model is centred on the cost of development. Mothers must provision their offspring until they are able to feed for themselves, that is, attain nutritional independence. The time spent in this developmental phase coupled with the energy expended will comprise the ‘cost’ of development. The minimum offspring size that allows individuals to reach nutritional independence must, then, increase with increasing cost of development.

As temperatures increase, developmental rate is expected to increase so that less time is spent in the developmental phase and metabolic rates (rates of energy use) are also expected to increase. The research team are suggesting that we consider how sensitive these two components are to temperature. If developmental rate is more sensitive to changes in temperature than metabolic rate, then the cost associated with provisioning offspring to achieve nutritional independence will decrease with increasing temperatures.

Or to put it another way, if developmental rate increases more than metabolic rate as temperatures rise, so that the developmental time is shorter in relation to metabolic rate, then the developmental cost is lower and offspring are smaller at higher temperatures. If, however, metabolic rate is more sensitive to changing temperatures than developmental rate then the converse is true; the developmental cost will increase with increasing temperatures and offspring are predicted to be larger at higher temperatures.

In order to develop and test these ideas the team needed to generate measures of temperature dependence of metabolic rate and developmental rate simultaneously, something that hadn’t been done before in a systematic fashion.

They started by methodically searching published literature to determine the relationship between the temperature that mothers experience and the size of their offspring. They then experimentally manipulated temperature to examine how developmental rate and metabolic rates changed in two very different species – the bryozoan Bugula neritinaand the zebrafish Danio rerio. They used the data from these experiments to develop the mathematical functions for their model to determine how the costs of development change with temperature. Finally they searched the literature again to get data on the temperature dependence of developmental and metabolic rates for a wide range of species because they wanted to test whether their model could apply more generally.

Amanda and her colleagues found that the offspring size / temperature relationship is widespread. Also in the two species they collected experimental data for, they found that development time is more sensitive to temperature than metabolic rates. This means that the overall costs of development decrease with temperature. What is more, they found that this pattern applies more broadly – for 72 species across five phyla the costs of development are higher at cooler temperatures.

Combining life history theory and metabolic theory has allowed the research team to provide a general explanation for offspring size / temperature relationships. In colder temperatures mothers show an adaptive response whereby they offset the increased costs of development by making larger offspring that possess greater energy reserves.

This research is published in the journal Ecology Letters.

This figure shows the relationship between two biological rates; development time (D) and metabolic rate (MR). In the left-hand graph we can see that as temperature increases from T4 to T1, development time is expected to decrease and metabolic rate is expected to increase. The right-hand panels demonstrate what is predicted to happen when b) the developmental rate is more sensitive to temperature than metabolic rate and so c) the total costs of development (and therefore offspring size) should decrease with increasing temperature. In d) the converse is true – metabolic rate is more sensitive to temperature than developmental rate and so e) total costs of development (and offspring size) will increase with increasing temperature.

New projects 2019

A number of large new projects will be getting underway in 2019 as a result of ARC funding schemes. Dustin Marshall and Matt Hall are now Future Fellows and Giulia Ghedini has received a Discovery Early Career Researcher Award. Dustin and Giulia will be using marine invertebrates to look into impacts of global warming whilst Matt is tackling the importance of sex in the evolution of infectious disease.

Within a given species, often the greatest heterogeneity that a pathogen will encounter will be due to differences between males and females. Yet, up until recently, insight into this crucial topic was driven by research into one sex, typically males.


Matt’s recent work has shown that, in the water-flea Daphnia magna, not only is pathogen fitness lower in males, but so is a pathogen’s evolutionary potential. What is more, the relative proportion of males in a population can fundamentally alter the overall transmission potential of a pathogen.

A. The graph on the left demonstrates that pathogen fitness is sex specific; in this case pathogen fitness is greater in females. B. The graph on the right indicates how changes in the relative proportion of males can increase the burden of disease for every individual.

This project was stimulated by Matt’s recognition that there is an absence of theory that explicitly considers how males and females can impact on the evolution and epidemiology of infectious disease.  Matt is seeking to address this imbalance and integrate sex-specific effects into a general framework for disease evolution and epidemiology.

Matt will be using the water-flea Daphnia magnaand its associated pathogens to provide an experimental system in which he can manipulate infections in males and females, characterise the degree of differentiation, and generate predictive models.


Dustin will be investigating how temperature affects the life-history stages of feeding and non-feeding larvae. Marine life histories show strong biogeographic patterns: warmer waters favour species with feeding larvae and cooler waters favour species with non-feeding larvae. Warming could be particularly problematic for Australian species because in 2012, Dustin discovered that Australian coastal species predominantly have non-feeding larvae. This means that future temperatures increases could affect native Australian invertebrates disproportionately relative to other regions of the world. (Put in schematic from application here)

Schematic of the data pipeline to estimate developmental energetics across temperature regimes. For each species, parents will be exposed to a range of temperatures, after which the size and number of offspring that are produced will be measured. These offspring will then have every phase of their energy usage and acquisition, from fertilisation through to metamorphosis estimated across an orthogonal temperature range. Dustin will then integrate these estimates to create a thermal energy performance curve for each species and use these data to parameterise models of connectivity, viability and life history evolution.

At the end of an intensive experimental period, Dustin will have quantitative estimates of how temperature alters the success of a range of species from the gamete to the juvenile. At this stage Dustin will work with collaborators to generate predictive models to determine

  1. how does temperature alter the relative advantages for each of the two developmental modes?
  2. how does temperature affect dispersal and connectivity among populations for each developmental mode? and finally
  3. how does temperature affect the distribution of marine organisms with feeding or non-feeding larvae?


Giulia will be investigating how global warming will affect entire ecological communities.

We already know that warming can affect individuals by reducing their body size and speeding up energy use, as well as reducing water viscosity.  But what we don’t know is how these changes at the individual level might play out at the population and community level and affect the energy intake or expenditure of whole communities.

Giulia will be looking at how changes in body size at the individual level interact with population and community level ecosystem functions.

Giulia is particularly interested in this knowledge gap and will be investigating the implications of warming sea temperatures for important ecosystem functions such as productivity, food web stability or resistance to invasion.

Giulia has planned a series of experiments, using communities of easily manipulated, sessile, marine invertebrates, to explore 4 main questions.

  1. How do changes in community size-structure and composition under warming alter the energy intake (phytoplankton) and expenditure (oxygen) of marine invertebrate communities?
  2. Since the availability of energy can mediate biological invasions, does warming alter the energy usage of communities so that they are more susceptible to invasive species?
  3. Are the responses of invertebrate communities to warming mediated by changes in their food (phytoplankton)?
  4. Given that warming reduces water viscosity, how does this mechanical effect alter food consumption and metabolic expenditure in marine communities of different size-structure?

Debating growth and reproduction

In a recent post we described a paper written by Dustin Marshall and Craig White and published in Trends in Ecology and Evolution (TREE). The published article was called “Have we outgrown the existing models of growth?” In it, Dustin and Craig suggest that the growth dynamics that biologists have long sought to understand probably emerge simply from hyperallometric scaling of reproduction.

Daniel Pauly is a fisheries scientist from the University of British Columbia and is a proponent of the Gill-Oxygen Limitation Theory (GOLT) of growth. This theory applies to water-breathing animals and is structured around the proposition that growth is necessarily constrained by the size of the gills and the oxygen they are able to extract from the water.

Professor Pauly argues in a letter to TREE that there is a good reason why growth is not considered to be influenced by reproduction in the context of GOLT. While he agrees that reproductive output tends to scale hyperallometrically in fish, he does not agree that fish slow their growth because they allocate increasingly more to reproduction. Instead, he thinks that as growth slows (due to oxygen limitation caused by physical constraints on gill size) increased allocation of resources is directed to reproduction.

In their rebuttal, Dustin and Craig summarise their difference in opinion as one of causality; while Professor Pauly argues that body size in fish is limited by gill area, they believe that organs evolve to provide capacity to meet an organisms requirements. Or, in other words, the trait of body size is the product of selection whereby the size of an organisms is the best it can be to maximise fitness in a particular environment. Most importantly, taken to its logical extension, Dustin and Craig argue that Pauly’s own arguments imply fish reproduction should decrease with size.

In a separate letter, Michael Kearney from the University of Melbourne suggests that a radical revision of growth models is premature. In this case, Associate Professor Kearney suggests that a mechanistic modelling approach (such as Dynamic Energy Budget theory) based on a thermodynamically explicit theory of metabolism is better suited to understanding growth than the phenomenological modelling approach proposed by Dustin and Craig.

While Assoc. Prof. Kearney argues that the Dynamic Energy Budget model can incorporate hyperallometric scaling by adjusting the ‘rules’ governing how much energy is allocated to reproduction, Craig and Dustin say that to do this requires a phenomenological approach and is an unjustified post hoc model fitting solution. According to Craig and Dustin, this means that Assoc. Prof. Kearney’s model is not strictly mechanistic, with some parameters estimated by fitting mechanistic functions and some parameters requiring empirical data (a phenomenological approach).

But there is some agreement. Dustin, Craig and Michael Kearney are all interested in seeing studies of growth and metabolism that are conducted in the context of a full accounting of energy and mass balances (food in, changes in length and weight, respiration, faeces and eggs out) to continue improving our understanding of why organisms are the size they are.

You can read the original article and the follow-up letters.

Why release small amounts of sperm slowly?

Sperm competition theory has been central to our understanding of male reproductive biology for many years and is dominated by the idea that males compete strongly to fertilise female’s eggs. But in many species the external environment will also influence reproductive strategies and, in their new publication, Colin Olito and Dustin Marshall ask an obvious but neglected question “what would reproductive strategies look like in the absence of sperm competition?”

Their interest was sparked by the fact that broadcast spawning species (e.g. seaweeds, corals annelid worms, sea stars and many fish taxa) release sperm and eggs to be fertilised externally, which provides an increased opportunity for the environment to influence the evolution of spawning strategies when compared to internal fertilisers.

In addition, broadcast spawners also have spawning strategies that differ markedly from predictions arising from classic sperm competition theory. For example, many broadcast spawning species have very long spawning times characterised by slow individual gamete release rates and, what is more, large males do not necessarily release more sperm than small males despite a large investment in gonads; neither strategy is predicted by classic theory.

Colin and Dustin devised two experiments to consider how fertilisation success changes with the amount of sperm released (ejaculate size) and the rate at which it is released. They used a marine intertidal polychaete worm, Galeolaria caespitosa, that has separate sexes and releases gametes initially into its tube and then, through rhythmic whole-body contractions, out of the tube in slow steady pulses.

A female Galeolaria removed from its tube and releasing eggs.

By repeatedly injecting different volumes of sperm (at the same concentration) and at different speeds into a flume set up to have laminar flow, Colin and Dustin were able to measure the fertilisation success of eggs placed ‘downstream’ of the sperm injection point.

Experimental set-up using the flume. Laminar flow was achieved by using collinators (drinking straws).

They used an experimental design that ensured that there was no variation in the number of males contributing to the pooled ejaculate used for the different experimental treatments. So, strictly speaking the experiments were not done in the absence of sperm competition, but, instead, in the absence of variation in sperm competition.

Colin and Dustin found that the benefits of releasing sperm quickly or slowly depended on ejaculate size: when only a small amount of sperm was released, it was better to release it slowly but when ejaculate size was larger and released at a faster rate, fertilisation success was greater for eggs further away. However, there was a substantial ‘cost’ associated with this higher fertilisation success for distant eggs. The more sperm males release, the more is wasted during sperm dispersal.

Colin and Dustin’s study suggests that slow sperm release rates are expected to evolve whether or not males experience strong sperm competition, and highlight the importance of taking account of selection from the external environment when seeking adaptive explanations for male broadcast spawning strategies.

This work has been published in the Journal of Evolutionary Biology.

The evolution of males and females depends on the environment

Opportunities for adaptation in females and males are mediated by life history and population characteristics that vary widely between species. Combining these factors with environmental heterogeneity can yield surprising evolutionary outcomes that are not always predicted by classic theories that deal with each factor separately.

Tim Connallon, Shefali Sharma and Colin Olito have analysed four simple models of evolution of female and male adaptations in changing environments. They compared the outcomes to classical population genetics models of sex-specific selection in stable environments and found some important differences.

Females and males make roughly equal genetic contributions to offspring. Consequently, the response to natural selection tends to depend equally on the pattern of selection in each sex. Selection does not necessarily increase adaptation of both sexes, but instead favours evolutionary changes in which the gains in adaptation for one sex are sufficient to offset any reductions in adaptation for the other. Such ‘sexually antagonistic’ selection is common and contributes to the maintenance of genetic variation.

Classical population genetics theory predicts that, where there are separate sexes, natural selection will favour genotypes that allow fitness across both males and females to be maximised. While these theories have proved extremely useful they tend to focus on evolution in constant environments – a condition that will be violated in many species.

Tim and his colleagues are interested in the outcomes of sex-specific selection in variable environments and how the life history and demography of a species’ can influence evolutionary dynamics. To this end they developed four mathematical models that vary in the life stage, or the sex, that disperses through a spatially or temporally heterogeneous environment.

In the first model, adults of both sexes disperse from the areas they were born in, with local selection happening before dispersal, and mating and reproduction happening after dispersal. This scenario applies to species with relatively immobile early-life stages, including many vertebrate and insect taxa. The second model considers taxa that have highly mobile early life-stages such as seed dispersal in plants or larval dispersal in many aquatic organisms. In the third example, adults from only one sex disperse from the area they were born in, prior to mating and reproduction. Sex biased migration is common in animals and can be strongly female biased or male biased. Finally, the fourth model deals with sex-specific selection that changes over time but is uniform across space.

When they ran these four different models and compared the outcomes with predictions from the classical population genetics models, they found the details of a species’ life history and demography were critical to determining the evolutionary dynamics of sex-specific adaptations.

For example, the models predict that conspicuous sex-limited colour polymorphisms (the simultaneous occurrence of multiple phenotypes limited to one sex only) should be particularly common in species that have strong sex-biased migration (scenario 3) and species where dispersal occurs early in the life cycle (scenario 2).

This work paves the way for diversifying the range of species that serve as models for studying sex-specific adaptations.

This work has been published in The American Naturalist.

Model outputs demonstrating that environmental variability promotes the maintenance of genetic variation beyond the comparatively restrictive conditions for polymorphism in a constant environment (solid grey versus dotted grey curves). Solid pink and blue lines indicate the likelihood of a beneficial allele becoming fixed for males and females in the different scenarios and over increasing selection pressure. Note that the pink curves overlay the blue curves in Models 1 and 4 and in Model 3 the equivalent results (with sexes reversed) would be obtained if there was female-limited dispersal.