While some geometric characteristics (length, surface area) of animals don't increase in direct proportion to any given mass increases, one would feel safe to assume that biochemical interactions (i.e. metabolic rate, MR) scale isometrically (10 times size=10 times metabolic rate). Mathematically expressed as MR=mass1, the exponent (1) indicates this direct proportionality (isometry). However, for individual organisms ranging in size from unicellular animals to mega-mammals, multiple studies have indicated that larger animals have lower than proportional (hypometric) metabolic rate increases with a consensus exponent of 0.75 – MR=mass0.75. While several (sometimes controversial) hypotheses have been formulated, we still lack clear agreement as to the underlying mechanisms of this ¾ power ‘law’. It does, however, suggest intricate cooperative interactions among cells, tissues and organs of individual organisms. Similarly, colonies of social organisms also show intricate interactions and non-centralized emergent properties. This raises the question of whether this level of social organization would extend to colony metabolic rate scaling comparable to that of individual organisms.
James Waters and colleagues, from Arizona State University, investigated the scaling relationships among 13 seed-harvester ant (Pogonomyrmex californicus) colonies ranging from 95 to 659 ants, all 10 months old. Using flow-through respirometry they measured metabolic rates of these fully functional lab-reared colonies. For comparison, Waters and colleagues then developed additive mathematical models to predict the collective metabolic rates, and the resultant scaling, of 13 hypothetical colonies, taking into account colony demographics such as numbers and masses of castes and developmental stages. They also measured metabolic rates of 20 groups of workers (1–225 ants) extracted from the lab colonies. In addition they determined individual sizes of all colonies, quantified worker activity in various sized colonies, and calculated the colonies' net growth rates.
The metabolic rate scaling for the additive models and also the extracted worker groups were isometric (MR=mass1), as expected, based on the direct proportional increases inherent in the additive models and from similar results obtained by previous authors working on isolated ant groups. The fully functional colonies, however, showed metabolic rate scaling indistinguishable from that of various sized individual organisms (MR=mass0.75)!
Waters and colleagues discuss a number of factors that could not explain this disparity, including colony composition and net growth rates. According to the team a greater proportion of ‘layabout’ ants in larger colonies only partly explained the reduction in the colonies' metabolic rate. The team showed that when activity is factored into the additive models, active individuals would need to increase their metabolic rate 25-fold to achieve the 0.75 scaling coefficient; however, Waters only measured a 6-fold increase between the ants' resting and running metabolic rates. Other factors that might contribute to the colony's reduced metabolic rate are reduced maintenance costs and/or metabolic rates that may vary relatively among colonies, and the fact that these colony size-independent metabolic rates may determine colony growth rates and effective sizes.
The team also speculates that individuals in insect colonies may integrate in a similar way to physically connected cells, tissues and organs in individual organisms and physically connected individuals such as sea squirts that show ¾ power scaling. Nevertheless, these findings clearly demonstrate that a colony of eusocial organisms behave metabolically like a single organism, possibly through networked behavioural interactions leading to patterns of resource and information transfer that influence metabolic scaling, and this lends more credence to the concept of the superorganism.
