Energy allocation explains how protozoan phenotypic traits change in response to temperature and resource supply

To survive and reproduce, living organisms need to maintain an efficient balance between energy intake and energy expenditure. Changes in environmental conditions can disrupt previously efficient energy allocation strategies, and organisms are required to change their behaviour, physiology, or morphology to cope with the new environment. However, how multiple phenotypic traits interact with one another and with environmental conditions to shape energy allocation remains poorly understood. To better understand this type of phenotype-environment interactions, we develop a predictive framework, grounded in energetic and biophysical principles that allows us to make predictions on how metabolic rate and movement speed should change in response to environmental temperature and resource supply, differentiating between short-term, acute exposure to novel conditions and longer-term exposure that allows acclimation or adaptation. We tested these predictions by exposing axenic populations of the ciliate Tetrahymena pyriformis to different combinations of temperature and resource availability. We measured population growth, cell size, respiration, and movement. Acute increases in temperature led to higher movement speeds and respiration rates, consistent with expectations from physical scaling relationships such as the Boltzmann–Arrhenius equation and the viscous drag acting on movement. However, by around 3.5 days after the introduction of Tetrahymena into a novel environment, all measured traits shifted toward values closer to those of the original environment. These changes likely reflect phenotypic acclimation responses that restored a more efficient energy allocation under the new conditions. Changes in cell size played a key role in this process by simultaneously affecting multiple phenotypic traits, including metabolic rate and the energetic costs of movement. In small microbial consumers like Tetrahymena, body size can change rapidly, relative to ecological and seasonal timescales. Changes in body size can therefore be effectively leveraged - alongside physiological and biochemical regulations -- to cope with environmental changes.

Methods
Cell cultures

Axenic cultures of Tetrahymena pyriformis strain 1630/1W were obtained from CCAP (www.ccap.ac.uk) and cultured at 20 degrees Celsius in a Proteose peptone - yeast extract medium 20 g Proteose peptone + 2.5 g yeast extract in 1L deionized water. Cells were maintained in the exponential growth phase at 20°C and 100% medium concentration for about 50 days (constant temperature, no light) in two replicates. Subsequently, cultures were split across 9 different treatments (3 temperature conditions x 3 medium concentrations) and adapted to the new conditions for at least three weeks (corresponding to a minimum of 20 generations), in four replicates for each condition.

Throughout the entire duration of the adaptation period, cultures were kept in relatively constant growing conditions by adopting a serial transfer regime, subculturing repeatedly into fresh growth medium. At the end of the adaptation period, cultures were further split across a wide range of temperature conditions (keeping the same medium concentration at which they were adapted) to characterise thermal response curves for different biological quantities (respiration, movement speed, and population growth rate). Specifically, we measured acute thermal responses for respiration and movement speed (from 1.5 to 3 minutes after cells were moved to the new temperature for speed recordings, and from approximately 30 minutes to three hours for respiration), and we measured long-term responses over a few days (2 to 6 generations, or 3.5 days on average, with differences depending on incubation temperature) for population growth, long-term movement speed, and cell volume changes.

Population growth
Assuming exponential growth throughout the experiment, the growth rate, expressed in the number of generations per day was calculated as log2(N(t) / N(0))/d where N(0) is the density at the beginning of a subculture, N(t) is the final density of the subculture, and d is the length of the subculture period in number of days.

Respiration
Oxygen consumption was measured using a 24-channel PreSens Sensor Dish Reader and 2ml oxygen sensor vials from PreSens https://www.presens.de/ inside thermal cabinets at the desired temperature. Population-level respiration rate was then converted to the rate of energy consumption per individual cell by accounting for population density in the vial and assuming an equivalence of 1mol O2 = 478576 J.

Imaging
Tetrahymena cultures were imaged under the microscope in cell-counting slides (BVS100 FastRead), which provide a constant depth of 100 um. A microscope camera (Lumenera Infinity 3-3UR https://www.lumenera.com/infinity3-3ur.html was used to record videos at 30 fps. The temperature while imaging under the microscope stage was controlled using a temperature-controlled stage (Linkam PE120 Peltier System).

Video tracking
We used custom-made software (available at https://github.com/pernafrost/Tetrahymena written in Python and based on the opencv, traktor, and scikit-image libraries to extract trajectories and measurements of cell size and morphology directly from the videos. The software returns measurements for each `tracked particle', where a particle generally corresponds to one individual cell. However, there isn't a perfect one-to-one correspondence between \textit{Tetrahymena} cells and particles: a cell moving in and out of the field of view of the microscope camera would be labelled as a different particle, and occasionally the identity of two cells swimming close to each other and overlapping in the images were swapped. Given the very large number of individual measurements both across and within experimental conditions, we ignored this small intrinsic potential for pseudo-replication in our analyses.

Movement analysis
The instantaneous movement speed of each tracked particle was measured from the trajectories as the displacement per unit time of the centre of mass of the particle, measured over a time scale of 1/10 of a second (3 frames). We took the median of all the instantaneous speed values along the trajectory as the individual speed of the particle. We excluded from the analysis particles that did not move and cells with an unnaturally small size that might have been incorrectly detected in the image. Unless otherwise specified, in the analyses and the figures presented here we focus on the top 20% faster-moving cells within each tested experimental condition. This is because fastest moving cells are more likely to be representative of the intrinsic limits of Tetrahymena swimming, while slowly moving cells might be stuck temporarily against the microscope slide or they might be moving at slow speed for reasons independent of their ability to move faster.

Estimation of cell volume
The video-tracking software fits an ellipse around each segmented Tetrahymena cell. Cell volume V was estimated based on the major and minor axes of the best fitting ellipse (respectively l and w) using the formula  V = 4/3 pi l w^2 / 8.