Metabolism, and metabolic rate tend to feature pretty highly in literature related to dinosaurs and other reptiles. For instance it is often stated that reptiles have metabolic rates around 1/10th those of similar sized mammals and birds, but what exactly does that mean? Talks of thermoregulation focus heavily on the role of metabolism, while allometric studies focus on how metabolism is affected by size. Given the prevalence of metabolic terminology in dinosaur and reptile papers/books, I thought it might be best to quickly give a review of metabolism, metabolic studies, and what all of that means for real animals.
Metabolism is everything
Metabolism is defined as the sum total energy expenditure of an organism. That is to say metabolism is the total energy an organism uses during its life. It is often broken up into the chemical reactions that build up resources (anabolism) and the reactions that break those resources down (catabolism). The amount of metabolism, or energy expenditure during a specific interval of time (seconds to days) is referred to as metabolic rate. From bacteria to blue whales, metabolism is the measure of all the energy that lets these critters go, and metabolic rates determine how much energy that is going to take. It can be measured in a variety of ways from respirometry to doubly labeled water and heart rate telemetry. The diversity of metabolic rate measurements is reflected in the units used to measure metabolism; which can range from watts/hour to milliliters of oxygen per minute, and even to joules per second.
Specificity is important
A key thing about metabolic rates is that they are plastic. They change depending on the situation presented. For instance one could measure the metabolic rate of a sleeping cat, and then compare it to measurements from that same cat while playing, or after eating a big meal. Metabolic rates ramp up when energy demand increases, and then ramp down when that energy demand decreases, or when the environment demands drastic energy cuts (e.g. starvation). Thus when measuring the metabolic rate of an animal it is important to decide exactly what kind of metabolic rate you are trying to measure.
And boy, oh boy are there a lot of different flavours to choose from.
One can measure: BMR, SMR, RMR, MMR, AMR, and FMR just for starters.
Those are a lot of initialisms, and they are just the most common ones. The choice of metabolic rate that one decides to measure is also going to dictate the technique that will be employed. So what do all these things stand for, and what technique is best for what? Let’s find out.
Technique 1: Respirometry
Respirometry is the most popular method for metabolic rate experiments. This is partly due to the cheap cost of materials, and also to the general relationship between the oxidation rates of adenosine triphopshate (ATP), and the substrates used to make them (i.e. carbohydrates, fats and protein). Respirometry is measured by obtaining the rate of oxygen consumption per mass per time (VO2/m/t). The mass can vary from grams to kilograms, and the time can be anywhere from seconds to hours (sometimes per day). Animals can be measured by either doing open flow respirometry — in which a mask is placed over the animal’s nose, or head, and oxygen is pushed through it at a known rate — or it can be measured using closed-circuit respirometry. This technique places the animal in an enclosed space with a known volume of air, allowing researchers to measure respiration by measuring the changes in that known volume of air.
Despite being the “gold standard” for MR measurements (Frapppell 2006), respirometry is not without its flaws. For instance the use of masks to control airflow, involves interaction with the animals and the placement of foreign (and often obstructive) parts on their bodies. This could result in handling stress and strange behaviours that only exist in the lab. Another, more important flaw with respirometry is the association with ATP production. While this association is practically constant over all animal groups studied so far, the difference between the oxidation of carbohydrates (which give the highest ATP yield) and protein (which gives the lowest) is approximately 10%. Since respirometry only measures the volume of O2 respired, it cannot tell the difference between the substrates. This means that conventional respirometry will always have a built in error bar that is +/- 10%. Sadly, this does not take into account general measurement error (a constant problem for any scientific experiment). Repeatability experiments with flow-through respirometry suggest that the actual error bars for even the most sophisticated flow-through respirometers are actually closer to 15-20% (Konarzewski et al 2005). So in fact, respirometry really gives us more of a “ballpark” figure for metabolic rate, than any concrete numbers.
Those caveats aside, even with the mush, respirometry has still proven to be a very effective way of measuring metabolic rate, and if one is using it in a comparative setting (say measuring aerobic scope, or comparing between taxa) the results can still hold weight.
Flow-through respirometry is generally the technique of choice for the following metabolic rate measurements:
BMR = Basal Metabolic Rate
This is a term reserved for automatic endotherms only. It is the lowest metabolic rate that can be attained by an automatic endotherm without dying. This is typically measured by fasting the animals for 72 hours, or more in a thermoneutral environment (i.e. one in which the animal does not need to actively work at to keep warm, or cool) and then measuring their VO2. BMR is tough to nail down for many automatic endotherms due to the rigour required to measure it (McNab 1997). Nonetheless it is an important measurement to take as it allows one to find the bare minimum energy levels required for that particular organism to live. These data can then be incorporated into body mass comparisons in order to create a predictive model of body mass and metabolism.
SMR = Standard Metabolic Rate
This is pretty much a carbon copy of BMR, but for bradymetabolic animals. The key difference is in the ways that bradymetabolic animals thermoregulate; especially under stress. When fasted, many if not all bradymetabolic animals will reduce their metabolism by becoming thermoconformers (i.e. they no longer regulate their temperature), or by actively seeking out temperatures that keep the metabolism running at a slower pace. Because of this lack of a temperature “set point,” one must determine the temperature in which they are going to measure SMR. Thus an accurate SMR will not only tell you the VO2 of the animal, but also the temperature with which it was taken (note: this should be true for BMR too, but it is less important and sometimes gets omitted). So, for instance, if one measured the SMR of an anole (Anolis) at 28°C and finds it to use 5 milliliters of oxygen per gram, per minute, then one would report it as 5 ml O2/min at 28°C. Once again I must reiterate: this is for fasting animals, and this represents the minimum energy required for them to live at that temperature. This means that while this information is helpful for more abstract problems, it should not be viewed as anything representing the “natural state” of the animal.
RMR = Routine / Resting Metabolic Rate
This is a more ecologically relevant measurement of metabolic rate. RMR is similar — but not equal — to BMR/SMR. This distinction is an important one that tends to get confused in many of the non-physiological literature (even dieting magazines seem to think the two synonymous). The difference is that now one is measuring the metabolism of a relaxed animal that is not undergoing physiological stress (i.e. fasting). RMR is often much higher than BMR/SMR. For instance consider the food requirements for an average person in the United States (according to the FDA). It is approximately between 1200-2000 calories per day. These calories are often spread out over 1-4 meals a day and do not account for rigorous daily activities. This is for a person who spends all day relaxing at home. Now consider a starving individual who is running closer to (or at) their BMR. This person might eat 1 meal a week, and can go as much as 3 weeks between meals (as exemplified by Mohandas Gandhi) all while running closer to 600-900 calories per day. The benefit of RMR is that it gives one an idea of what energy level a healthy animal idles at.
MMR = Maximal Metabolic Rate
This is a measure of how much energy an animal is using per unit time while undergoing some intense and strenuous activity such as the “frenzy” period of hatchling sea turtles as they head to the water, the flight response of a lizard that has been scared, or a racehorse during the Kentucky derby. MMR is a measure of the highest aerobically sustainable energy use an animal can perform. It is often many times higher than basal/standard, or resting metabolic rate (up to 100 times higher in insects [Davis & Fraenkel 1940]).
AMR = Active Metabolic Rate
This is similar to maximal metabolic rate, except that this measures RMR + daily activities such as locomotion, hunting, eating, thermoregulation etc. It can be many times the resting metabolic rate. Since AMR requires animals to move about and act natural, respirometry is harder to do. Closed-circuit systems have been used successfully, but are often limited to smaller taxa (though see Castellini et al 1992 for a cool exception). Because of the problems associated with getting accurate respirometry data, AMR measurements are best used with other methods such as…
Technique 2: Doubly Labeled Water (DLW)
DLW measures the CO2 concentrations within an animal by comparing the difference between deuterium (heavy hydrogen) and the 18-oxygen isotope (O18) over a period of time (Lifson & McClintock 1966). This is done by injecting water that has been laced with deuterium and O18. A blood sample is taken before, shortly after injection (a few minutes to a few hours depending on the size of the animal), and then a few days, to a few weeks later. Changes in deuterium concentration are directly related to water turnover via urination, defecation, and respiration. This is used as a control when comparing the changes in O18 concentration, as O18 while lost through water turnover, is also lost via conversion to CO2. This causes O18 to get “washed out” of the body faster than the deuterium. The difference between the O18 and deuterium wash out concentrations approximates the level of CO2 production in the animal’s body. The big advantage of DLW over respirometry is that it limits the time that researchers are interacting with their study animals. Other than a few brief moments where the animal must get pricked with a needle, the interactions are practically nonexistent, and the data that comes back represent the most ecologically relevant results available (since the animals being studied should be acting normally in their environments).
DLW has proven invaluable for FMR analyses (see below), and related isotope work has allowed researchers to even determine diets of free ranging animals without the need for invasive procedures like stomach pumping. However there is a caveat to all of this. DLW measurements are unbiased. The results obtained could include anything from relaxation (RMR), to active foraging (AMR), and predation/predator avoidance (MMR). They don’t account for any environmental factors, and really just provide one with a snapshot of the period of time in which the experiment took place. So unless the researchers are being particularly observant of an animal’s behaviour during the study period, DLW can only give one a mean energy expense for a study animal throughout that period of time. Other disadvantages include the high costs of the isotopes themselves, which precludes the obtainment of large sample sizes for big animals (from $1000 for a 70kg human [Butler et al 2004] to over $7500 for a 300kg leatherback sea turtle [Wallace et al 2005]). While DLW seems to have better precision than flow-through respirometry, with a built in error of approximately 6.5% (Speakman 2004), this error increases in animals where the ratio of produced CO2 to water production decreases, such as in marine bradymetabolic animals.
So what does DLW actually give us?
FMR = Field Metabolic Rate
In many ways field metabolic rates represent the most ecologically relevant metabolic measures out there. These are measurements of the metabolism of a taxon in its natural environment, doing what it naturally does. By comparing a series of FMR studies on animals from the same ecosystem it is possible to infer which species are more active in that environment, as well as food use by these different animals. However, since FMR studies do rely heavily on the doubly labeled water method, this means they are also subject to all the errors that are associated with that method. The biggest problem with FMR measurements is that while they may give a fascinating snapshot into the energetics of an animal in a specific environment during a specific period of time, the huge array of unaccounted variables in these measurements makes FMR studies very hard to reproduce (Nespolo and Franco 2007). Still, the potential for FMR to give a “real world” result for animal energetics in a particular environment remains attractive, and has resulted in some milestone publications on this very subject (e.g. Nagy et al 1999).
Technique 3: Heart Rate Telemetry
One way to handle the errors associated with doubly labeled water is to just avoid that technique entirely. Heart rate telemetry techniques seem underpublicized compared to DLW and respirometry, despite the 60 year pedigree (starting with Lundgren 1946). They operate under the assumption that a change in heart rate is a major component of the response of an animal to increase oxygen demand (i.e. increased metabolic rate). This technique also assumes that the oxygen pulse (OP) remains constant. Oxygen pulse is defined as the heart stroke volume multiplied by the O2 content of arterial blood minus the O2 content of mixed venous blood. If OP is constant, or shows some kind of systematic change, then there should be a linear relationship between volume of O2 consumed and heart rate. If this assumption holds true then one can theoretically determine respired O2 by simply measuring the heart rate. In order to measure this, animals must have a data logger implanted in them. This bit of surgery does count as a fairly extreme interaction, but it is one that happens only once. From there on out the data is transmitted wirelessly through standard telemetry practices. Depending on the technology in use at the time, data loggers can house daily heart rate data anywhere from a few weeks (Woakes et al 1995) to over a year (Butler et al 2004). The advantages of heart rate telemetry over DLW methods is that they can actually give FMR data that is more precise, rather than just an average. For instance an increase in heart rate would correspond to an increase in metabolism, which tells one that RMR is no longer being measured. Further data logger technology now allows for the recording of other variables as well, allowing the correlation of heart rate with other ecologically relevant things (e.g. temperature). There is of course a downside to all of this. The underlying assumptions (i.e. that heart rate and oxygen consumption are constant) may not be true. In fact we know that this is not true just from general data on humans. A person who has been training to run a marathon is going to have a slower beating heart than a weightlifter, who will have a slower beating heart than a couch potato. The buildup of stroke volume will skew the relationship between OP and heart rate, resulting in lower FMRs than in reality. Further, since the cardiovascular physiology of taxa vary so greatly across Animalia, a separate calibration equation must be derived for each new species (Butler et al 2004), as well as for any taxa that show unexpected results (e.g. exercise trained animals). So despite the greater precision of this technique, the amount of rigour required to stay on top of things poses a greater barrier to entry than the (slightly) easier doubly labeled water method.
All of these just brush the surface of metabolic studies. The initialisms we’ve encountered here are but the tip of the metabolic iceberg. Each one of them can have specifiers related to digestion, locomotion, thermoregulation etc. The three techniques I have outlined above are just the most popular. There are other variants, and even more rigorous techniques (e.g. the use of structural equation modeling or factor analysis), but these techniques appear to pose an even greater barrier to entry such that they rarely are employed (Nespolo & Franco 2007).
I hope that this brief (well, I was trying to make it brief) intro into metabolic rate studies helps drive home the point that not all metabolic rates are equal. When reading any literature that gives a comparison between the metabolic rates of different taxa, it’s important to question what the criteria were for the study and what metabolic rates were being measured.
Bringing us back to the original statement at the beginning of this post: it is often said that reptiles have a metabolism that is 1/10th that of similar sized mammals, or birds. Well, is this a resting metabolism, or a standard metabolism? If the latter, then what was the criteria for these measurements? What techniques were used, and what temperatures were the animals being housed in? Were the temperatures equivalent to similar BMR studies in those mammals? All of these are important questions to ask if one wants to get an accurate comparison between disparate taxa, and especially if one is going to argue for one metabolic life history over another.
As the title suggests, there is a second installment of this post in the works. I left one other major metabolic rate measurement out, that of mass-specific metabolism. Given the importance often placed on this, and the controversy surrounding it, I thought it best to give it its own post.