What role does oxygen supply play in determining a species’ thermal tolerance?

It is generally recognised that animals perform best at certain temperatures, so-called optimal temperatures. To understand how measures of performance, such as growth or running ability, will change with changing temperatures we need to understand the physiological processes limiting performance.

One compelling theory, known as the oxygen and capacity-limited thermal tolerance (OCLTT), suggests that a reduction in oxygen availability limits performance.

As temperature shifts away from optimal, the demand for oxygen in tissues is thought to outpace the rate of supply. This means a shift to anaerobic metabolism, a process far less efficient than aerobic metabolism, causing a reduction in performance.

Emily Lombardi, Candice Bywater and Craig White wanted to test the theory and used the speckled cockroach as their experimental animal because they are easy to breed and keep in the lab. Unlike other studies testing the OCLTT hypothesis, their interest was in the less extreme ends of the temperature range, which the OCLTT hypothesis specifically addresses.

Emily, Candice and Craig predicted that performance would be reduced (↓) in hypoxic environments and sub-optimal temperatures but increased (↑) in hyperoxic environments at optimal temperatures and stay the same (−) as atmospheric oxygen conditions at optimal temperatures when in a hyperoxic environment but at sub-optimal temperatures.

First, Emily and her colleagues calculated the temperature that optimised growth by allowing juvenile cockroaches to develop for 35 days at 8 temperatures ranging between 10 and 36 °C. They also determined the temperatures (both above and below the optimal temperature) where growth was reduced by 32%. These 3 temperatures (optimal, lower and upper developmental temperatures) were then used in oxygen manipulation experiments.

Next, they ran an experiment including the three temperatures plus three oxygen concentrations; atmospheric (21%), hypoxic (10%) and hyperoxic (40%). Cockroaches were assigned to one of the 9 treatments and growth rate was measured repeatedly over 5 weeks. Running performance – time on a treadmill – was measured at the end of the 5 weeks at the three temperatures. Tracheal morphology was also quantified because, in some species, changes to the oxygen delivery system can alleviate the demand for increased oxygen supply.

If the OCLTT theory held, then the team expected to see a difference in performance at the different oxygen concentrations at each temperature. But, instead, they found increasing oxygen concentrations did not mitigate the effects of sub-optimal temperatures. There was also no evidence that tracheal morphology changed as a result of the developmental temperature or oxygen environment. The team concluded that oxygen supply was not the main determinant of temperature-related performance limitations for the speckled cockroach.

It seems, the OCLTT may not provide the unifying theory its proponents hoped, but instead, a species’ thermal tolerance is likely shaped by a range of factors.

In fact, they found that growth did not differ between oxygen concentrations at the lower temperature, but growth was lower in hypoxic environments at optimal temperatures when compared to atmospheric oxygen concentrations and lastly growth was lower in hyperoxic environments at high temperatures compared to atmospheric oxygen concentrations. All points that share a letter are not significantly different from one another.

This research was published in the Journal of Experimental Biology.

Competition plays a part in maintaining variation in metabolic rates

We know that the rate at which organisms use energy (metabolic rate) varies substantially between individuals of the same species, even after accounting for size and temperature. What we are less sure about, is, why we see this variation.

When Amanda Pettersen and her colleagues thought about this question they considered it plausible that the competitive environments that individuals find themselves in might be important in determining whether a faster or slower metabolic rate is selected for. They wanted to find out whether variation in competition affects selection on metabolic rates and whether that could account for the variation in metabolic rate that persists more generally.

To test this idea Amanda used the model species Bugula neritina because it allowed the team to collect larvae in the lab, measure size and metabolic rate of larvae before assigning the larvae to one of three ‘competition’ treatments. Once the larval measurements were made, larvae were randomly assigned to either a no-competition environment, a competitive environment where they were put with other Bugula neritina (intraspecific competition) or, a competitive environment with an established mixed-species community (interspecific competition).

Individuals were then returned to the field and monitored weekly for survival, growth, age when first become reproductive and fecundity (total reproductive output).

Following measurements of size and metabolic rate, larvae were settled on small plates that were either kept bare (no competition), had other Bugula on the plate (intraspecific competition) or had an established community already there (interspecific competition). Plates were submerged in the field and monitored weekly.

The team already knew that higher metabolic rates are linked to faster growth, earlier onset of reproduction, and a shorter lifespan, while low metabolic rates are associated with a slow pace-of-life (slow growth, late onset of reproduction and long lifespan).

This experiment showed that individuals with higher metabolic rates were more likely to survive, more likely to reproduce and had greater numbers of ovicells at the start of reproduction, in the more intense, interspecific competition treatment. The team speculates that the individuals with higher metabolic rates grow more quickly enabling them to reach resources such as food and oxygen that are less available to smaller, slower growing organisms. There is a downside: these higher metabolic rate individuals also had a shorter lifespan.

The expectation amongst evolutionary biologists is that where there is strong selection pressure for a particular trait then the variation within that trait is reduced. So even accounting for the shorter lifespan, the overall increase in reproduction should mean that a higher metabolic rate is strongly selected for in competitive environments.

But the team also found that individuals with lower metabolic rates had a higher probability of living for longer in the absence of competition and would have continued to reproduce long after the individuals with a higher metabolic rate had died.

Bugula live on hard substrates and these areas commonly vary in their composition. Complex three-dimensional, highly diverse communities are interspersed with sparse populations of a few species and patches of bare areas. Individuals with lower metabolic rates that happen to arrive at a bare patch will have a long reproductive period and high overall fecundity. In contrast, individuals with a high metabolic rate that settle in amongst other species will be reproductive quickly and so successfully produce large numbers of offspring.

Because Bugula larvae are likely to find themselves in a variety of different competitive environments – even when they settle relatively close to each other – there is a strong evolutionary argument to explain the persistence of variation in metabolic rate, for Bugula at least.

This research is published in the journal Evolution Letters.

Metabolic rate, context-dependent selection, and the competition-colonization trade-off

Authors: Amanda K Pettersen, Matthew D Hall, Craig R White, and Dustin J Marshall

Published in: Evolution Letters


Metabolism is linked with the pace-of-life, co-varying with survival, growth, and reproduction. Metabolic rates should therefore be under strong selection and, if heritable, become less variable over time. Yet intraspecific variation in metabolic rates is ubiquitous, even after accounting for body mass and temperature.

Theory predicts variable selection maintains trait variation, but field estimates of how selection on metabolism varies are rare.

We use a model marine invertebrate to estimate selection on metabolic rates in the wild under different competitive environments.

Fitness landscapes varied among environments separated by a few centimetres: interspecific competition selected for higher metabolism, and a faster pace‐of‐life, relative to competition‐free environments.

Populations experience a mosaic of competitive regimes; we find metabolism mediates a competition-colonization trade-off across these regimes. Although high metabolic phenotypes possess greater competitive ability, in the absence of competitors, low metabolic phenotypes are better colonizers.

Spatial heterogeneity and the variable selection on metabolic rates that it generates is likely to maintain variation in metabolic rate, despite strong selection in any single environment.

Pettersen AK, Hall MD, White CR, Marshall DJ (2020) Metabolic rate, context-dependent selection, and the competition-colonization trade-off. Evolution Letters PDF DOI

Lab life during lockdown

Lab life has, of necessity, been curtailed throughout the world but it has provided an opportunity for researchers to spend time trawling the literature for data to use in meta-analyses. Our lab is no different and so in this edition of lab life we aim to give an overview about what some of our members have been working on.

Michaela Parascandalo has continued the data mining of reproduction data to enable the group to ask: do larger individuals produce disproportionately more gametes / offspring than smaller individuals in taxa other than fish? So far, we have more than 1000 species across 10 phyla and more than 100,000 data points.

One tricky element of this data collection has been converting the different body size measurements to mass. For example, a paper may present data on head width in wasps, hind tibia length in grasshoppers, or carapace lengths in crabs but Michaela needs to convert this to a measure of mass. To do this she has had to create a repository of morphometric allometries for a number of different species, another great resource for other meta analyses that the lab group might do in which body mass needs to be calculated.

A separate study is also underway that is looking at not only the number of offspring but the size of those offspring in relation to maternal size. So, Melanie Lovass, with help from Michaela, has been compiling data to enable Hayley Cameron and our (now virtual) visitor Darren Johnson from California State University to ask the question: do bigger mothers produce bigger and/or more offspring and are there any differences between warm-blooded and cold-blooded animals?

PhD student, Emily Richardson is particularly interested in organisms that have complex life histories or, in other words, go through metamorphosis to become adults. Emily is gathering data on growth rates in amphibians, fish and marine invertebrates to test the theory that growth rate is maximised relative to mortality rate at the time of metamorphosis, which would mean that fitness is increased.

George Jarvis and Sam Ginther are also doing meta-analyses that relate to their PhD projects. George, like Emily, is interested in organisms with complex life histories but he is looking at large scale evolutionary change in metabolic rate. For his meta-analysis he is compiling metabolic data from marine invertebrates and looking at how metabolic rates vary between species with different developmental modes. With this work, he hopes to better understand the evolution of metabolic rate in complex life cycles.

Sam is interested in the cost of reproduction. He is collecting data on metabolic rates in reproductive and non-reproductive adults as well as their offspring.  This will help him understand how the energy used for reproduction affects the production of offspring in species with dramatically different life histories.

Louise Noergaard has just started a post doc in the CGB and is busy working on a collaborative project between Dustin Marshall and Beth McGraw from Penn State University. Louise is looking at the relationship between wing length and body mass in mosquitoes and assessing how these measures of size relate to lifetime reproductive output. This information can then be put into a model that will consider how these measures of size and reproductive output affect existing predictions of mosquito spread.

Several members of the group are interested in organisms with complex lifestyles. The marine tubeworm Spirorbidae shown here is an example. Larvae are released from the adult brood chamber via a split in the chamber wall (a), the larvae are non-feeding, free swimming (b), when they find a suitable surface the larvae settle and start to metamorphose (c/d) and once metamorphosis is complete newly settled juveniles start to feed.

The effect of ambient oxygen on the thermal performance of a cockroach, Nauphoeta cinerea

Authors: Emily J Lombardi, Candice L Bywater, and Craig R White

Published in: Journal of Experimental Biology


The oxygen and capacity-limited thermal tolerance (OCLTT) hypothesis proposes that the thermal tolerance of an animal is shaped by its capacity to deliver oxygen in relation to oxygen demand. Studies testing this hypothesis have largely focused on measuring short-term performance responses in animals under acute exposure to critical thermal maximums. The OCLTT hypothesis, however, emphasises the importance of sustained animal performance over acute tolerance.

The present study tested the effect of chronic hypoxia and hyperoxia during development on moderate to long-term performance indicators at temperatures spanning the optimal temperature for growth in the speckled cockroach, Nauphoeta cinerea.

In contrast to the predictions of the OCLTT hypothesis, development under hypoxia did not significantly reduce growth rate or running performance, and development under hyperoxia did not significantly increase growth rate or running performance. The effects of developmental temperature and oxygen on tracheal morphology and metabolic rate were also not consistent with OCLTT predictions, suggesting that oxygen delivery capacity is not the primary driver shaping thermal tolerance in this species.

Collectively, these findings suggest that the OCLTT hypothesis does not explain moderate to long-term thermal performance in N. cinerea, which raises further questions about the generality of the hypothesis.

Lombardi EJ, Bywater CL, White CR (2020) The effect of ambient oxygen on the thermal performance of a cockroach, Nauphoeta cinerea. Journal of Experimental Biology PDF DOI

You are what you eat, but does it matter when you eat it?

We all know the expression “you are what you eat” but do juvenile diets determine your adult size or can adult diets change things? Gonçalo Poças, Alexander Crosbie and Christen Mirth have used a model organism – the fruit fly Drosophila melanogaster – to explore this question.

They found that while larval nutritional conditions play the dominant role in determining both adult body weight and appendage size in D. melanogaster, the adult diet can adjust body weight as the flies age.

The team manipulated both the quality and energy quantity of diets that larval and adult flies received in two experiments.

In the first experiment calorie content was held stable but the ratio of protein to carbohydrate was either 1:2 (high quality) or 1:10 (low quality). In the second experiment flies were fed a high quality diet but with two different calorie contents. The low calorie content diet was 25% of the high calorie diet.

Importantly, larvae were reared on one of the two diets in each experiment and then after “eclosing” (emerging as an adult) they either remained on the same diet or were switched to the alternative diet. This allowed the researchers to tease apart at what life-stage diet was most influential in determining measures of adult size.

Adult weights, wing area and femur size were measured at 3, 10 and 17 days old. The team predicted that adult weight would be more influenced by adult diet but that wing area and femur size would be determined by larval diet.

Schematic of the experimental design. Larvae were fed alternative diets in two experiments that looked at the quality of the diet or the energy content of the diet. Once flies emerged as adults they were kept on the same diet or switched to the other diet and measures of size were taken at 3, 10 and 17 days.

For the most part, the team found that the larval diet contributed more to differences in adult weight, wing area, and femur size in both males and females, but the quality of the larval diet had greater effects on adult size than the calorie content.

In addition to the effects of larval diet on adult size traits, Gonçalo and his colleagues found that animals subjected to poor larval nutrition were able to increase their body weight if maintained on good quality diets during the adult stages.

One surprising result the team uncovered was that the adult femur size changed with age, and depended on both the larval and adult diet. This suggests that the size of adult appendages might not be as fixed as previously thought.

So, to return to our initial question. Yes, the juvenile diet is very important in determining measures of adult size, but adult diets can make up for nutritional deficiencies in earlier life to some extent – in fruit flies at least.

This research is currently in press in the Journal of Insect Physiology.

Can communities minimise energy wastage over time?

Understanding how energy is used (or not used) in individuals, populations and communities has fascinated biologists for many years. Giulia Ghedini, an Australian Research Council research fellow in the Centre for Geometric Biology, is particularly interested in how communities use energy. Giulia and her colleagues have found that older communities of marine phytoplankton waste less energy than new or early successional communities.

In a previous post, we described how Giulia and her colleagues used marine invertebrates to test whether ‘older’ communities minimised energy wastage as predicted by Robert MacArthur in 1969. MacArthur basically said that, due to competition for resources, a community of species will, over time, maximise the efficiency with which it consumes resources meaning that older communities have very little energy (resource) wastage.

If MacArthur was right, then we can better understand and predict how disturbances that change the age or successional stage of a community might impact community function and invasion risk.  An older community might be able to sustain a greater biomass with the same resource requirements of a younger community. Not only that, but a community that efficiently uses all the energy or resources available makes it difficult for a new species to gain a foothold.

While the team found some support for MacArthur’s theory in their initial study, it was far from clear-cut and they thought that this might be because their experimental setup allowed for predation and immigration; MacArthur’s theory relates to a closed system where competition between species is the driving force for community structure and function.

To tackle this problem Giulia and her colleagues designed a new experimental setup using a completely different system: marine phytoplankton. They were able to create ‘starter’ assemblages of phytoplankton in the lab. They used six species and then followed these assemblages through time measuring energy inputs (light and nutrients) and energy use (photosynthesis, metabolism and overall production). This allowed them to work out how much energy from resources was not used and how much was lost to maintenance  – metabolism and mortality – the two components of MacArthur’s ‘wastage’. They also tracked changes in species composition and the size structure of the communities to understand their links with energy wastage.

The team found some marked changes in the phytoplankton communities. Larger species became more dominant over time and this was consistent in all 20 of the experimental communities. Surprisingly, neither changes in dominant species nor the homogenisation of communities were driving changes in energy flux.

So, while the composition of the species changed in a consistent way as the communities ‘aged’, the functioning of communities tracked less consistently, meaning that the oldest communities were not necessarily the most energy efficient. But there was a general pattern that fitted with MacArthur’s prediction that older more established communities would minimise energy wastage. And, as predicted, the changes were due to the two components of energy use changing in opposite directions.

The combined results from the team’s work suggest that overall patterns in community function might be predictable over time and that MacArthur’s minimisation principle might apply across very different systems.

Tracking species composition and energy use through time enabled Giulia and her colleagues to see that as the community was better able to harvest and use the available resources so that resource waste decreased over time (green), more energy was lost to metabolism and death (magenta). The black line represents the community trade off between the ability to uptake and use resources and the costs of maintenance that this competitive ability requires.

This research was published in the journal Ecology.

When does diet matter? The roles of larval and adult nutrition in regulating adult size traits

Authors: Gonçalo M Poças, Alexander E Crosbie, and Christen K Mirth

In press in: Journal of Insect Physiology (preprint)


Adult body size is determined by the quality and quantity of nutrients available to animals. In insects, nutrition affects adult size primarily during the nymphal or larval stages. However, measures of adult size like body weight are likely to also change with adult nutrition.

In this study, we sought to the roles of nutrition throughout the life cycle on adult body weight and the size of two appendages, the wing and the femur, in the fruit fly Drosophila melanogaster.

We manipulated nutrition in two ways: by varying the protein to carbohydrate content of the diet, called macronutrient restriction, and by changing the caloric density of the diet, termed caloric restriction. We employed a fully factorial design to manipulate both the larval and adult diets for both diet types.

We found that manipulating the larval diet had greater impacts on all measures of adult size. Further, macronutrient restriction was more detrimental to adult size than caloric restriction. For adult body weight, a rich adult diet mitigated the negative effects of poor larval nutrition for both types of diets. In contrast, small wing and femur size caused by poor larval diet could not be increased with the adult diet.

Taken together, these results suggest that appendage size is fixed by the larval diet, while those related to body composition remain sensitive to adult diet. Further, our studies provide a foundation for understanding how the nutritional environment of juveniles affects how adults respond to diet.

Poças GM, Crosbie AE, Mirth CK (2019) When does diet matter? The roles of larval and adult nutrition in regulating adult size traits. Journal of Insect Physiology PDF DOI


Are there any advantages to being smaller in higher temperatures?

Biologists have been familiar with a pattern of smaller body sizes with increasing temperatures for a long time, in fact, so familiar that Bergmann dubbed a “Temperature-Size Rule” in 1847.

Like many things to do with size, it is difficult to separate the effects of temperature on size from other traits that co-vary with size; metabolism for example. It may be that higher temperatures cause the evolution of faster metabolic rates and metabolic rate is genetically correlated with size. So that it is, in fact, metabolic rate that is the target of selection, not size.

Martino Malerba and Dustin Marshall were again able to take advantage of the evolved large and small algal cells to see if they could unambiguously assign any effects of temperature on size, to size alone. They wanted to find out if (and how) temperature affected fitness for different sized organisms.

To do this they used algal cells that had experienced 290 generations of artificial selection and where large selected cells were 13 times bigger than small selected cells. They then exposed these different lines (including the control lines) to three temperatures 18 °C, 22 °C and 26 °C and measured cell size, population density and cell production rates after three and six days.

They found that the smaller cells did better at higher temperatures; that is, the fitness proxies of cell production rate and population densities were both greater for small cells at higher temperatures. This means that Martino and Dustin have shown that size on its own can affect performance across different temperatures.

They then wanted to know why are cells smaller at higher temperatures; what is the advantage? It has long been thought that smaller cells do better in warmer temperatures because they have a greater surface-area to volume ratio. This would make them better able to take up resources such as nutrients, CO2 and light at the same time as increasing temperatures increase a cell’s demand for resources through increased enzyme activity and protein synthesis.

If this was the case, Martino and Dustin expected the large and small cells to show differences in performance at higher temperatures when resources were abundant (days 0 to 3) compared to when resources were depleted (days 3 to 6). But they found no difference in the fitness of large and small cells that related to resources suggesting that advantages of smaller cells at higher temperatures was not related to a greater surface-area to volume ratio.

Instead they measured the concentrations of reactive oxygen species in their selected lines of large and small cells. Reactive oxygen species are known to increase oxidative stress, damage DNA and so reduce the performance of a cell and also accumulate at higher temperatures. Martino found that the larger cells had almost five times more reactive oxygen species than smaller cells. And the larger cells had relatively smaller nuclei, meaning that there was twice the reactive oxygen species loading around the nuclei in large selected cells.

Martino and Dustin think that it is likely that small cells do better at higher temperatures, not because they are able to access more resources per unit volume, but because they are less prone to toxicity from reactive oxygen species.

This research was published in the journal Evolution.

Effects of two years of artificial size-selection (290 generations) on the cell volume of a phytoplankton species (Dunaliella tertiolecta).
Fitness increased with temperature (positive slope) in small-selected algae and decreased with temperature (negative slope) in large-selected algae. This finding confirms that reducing the size of a species automatically provides a fitness benefit at warmer temperatures.

Water temperatures and viscosity will both change with ocean warming but how will they affect male fertility in the tubeworm Galeolaria?

Global warming will increase ocean temperatures at the same time as it reduces seawater viscosity and Evatt Chirgwin wanted to know how this combination of physiological and physical change would affect male fertility in a small tubeworm. He found that both these factors independently reduced male fertility, and together altered selection pressures on sperm morphology.

Most marine species release gametes into the water column and successful fertilisation depends on a sperm locating and fusing with an egg. This high-risk strategy is in stark contrast with many terrestrial species where sperm and eggs interact in the controlled environment of a female reproductive tract, making marine species more vulnerable to global warming.

Projected ocean temperature increases are expected to reduce male fertility because exposure to temperatures outside the usual range can disrupt physiological processes and cell function. But the viscosity or ‘thickness’ of the seawater will also change with increasing temperatures, and Evatt was interested in understanding how the fact that sperm are able to move more easily through the water would affect male fertility.

Because these two things tend to change together, up until now no one has considered how decreases in viscosity at higher temperatures might alter fertility as well as selection pressure on sperm structure. Sperm with larger heads have increased ‘drag’ while a long tail can increase swimming speeds – these might not matter so much when seawater is easier to pass through.

So how do temperature and viscosity affect male fertility and the selection forces acting on the size and shape of sperm? Evatt and his supervisors (Keyne Monro and Dustin Marshall) measured fertilisation success at three temperatures and used a hydrophilic polymer that allows warmer water to be adjusted to the same viscosity as cooler water (but not the other way around).

Evatt measured head size, midpiece size and tail length in the sperm of 157 males that had access to eggs from a variety of females in five different fertilisation environments.

Evatt manipulated both temperature and viscosity to give five fertilisation environments for the 157 males that he had measured head size, midpiece size and tail length in the sperm. Sperm from each male added to eggs from a number of different females in each of the five environments and fertilisation success recorded.

The team found that the isolated effects of temperature and viscosity each caused fertility to decline by around 5% from current to moderate warming and by another 5% from moderate to extreme warming. But temperature and viscosity acted together to alter selection on sperm morphology. The ‘midpiece’ that houses the mitochondria, was a target for selection at the projected, warmer environments. A shorter midpiece was favoured in moderate warming environments, while a wider midpiece was favoured at the more extreme, longer-term projections of warming.

Sperm midpiece size was a target of selection with both temperature and viscosity acting in combination. Narrower midpieces are favoured by current conditions, shorter midpieces are favoured by projected warming environments and more extreme or longer term warming favoured wider midpieces.

Evatt and his supervisors think that since the midpiece contain the mitochondria that provide energy, it is probable that changes in temperature and viscosity will change the energy requirements of sperm during the location and fertilisation of eggs.

For the first time, the team show how projected changes in water temperature and viscosity may impact the fertility of marine populations and expose sperm to novel evolutionary pressures that may drive them to adapt in response.

This research was published in the journal Functional Ecology.