INTRODUCTION 

Nutrition is an upstream organizer of most organismal processes, and thus dealing with variation in the availability of nutrients is among the most consequential challenges, and selective agents, faced by all organisms (Stearns, 1992; Zera & Harshman, 2001). While many nutrients are necessary for maintenance and growth, the majority consumed and used are macronutrients ingested in high quantity, such as proteins, carbohydrates, and fats (Raubenheimer, Simpson, & Mayntz, 2009). The molecular mechanisms that translate incoming macronutrients into maintenance, new tissue, or storage, are well understood, and tend to be highly conserved (Conlon & Raff, 1999).
Superorganisms, like eusocial insects, are groups of individuals of the same species operating in a synergistic way (D. S. Wilson & Sober, 1989). Superorganisms are characterized by the reproductive division of labour, such that some individuals do not reproduce (somatic – germ line division) (Hölldobler & Wilson, 2009). The nutritional challenges of any cell or organism are also faced by superorganisms, but are compounded by the additional level of complexity (cooperating organisms on top of cooperating cells, etc.) – the labour that is divided amongst cells in ‘unitary’ organisms is divided amongst individuals in the superorganism. How, then, do nutritional challenges scale from organism to superorganism? Do organism and superorganism have different nutritional optima and/or face the same nutrition-mediated trade-offs? If individual and superorganism optima are not aligned, then processes occurring at the individual level may limit those at the superorganismal level.
Social insect colonies undergo a coordinated development that is responsive to the environment and also lineage/species specific, termed ‘sociogenesis’ (E. O. Wilson, 1985). ‘Sociometry’ is a general term applied to the metrics associated with superorganisms, such as colony size, individual size (and size distributions), the numbers and ratios of different castes, among others. As a general pattern in studies of sociometry and sociogenesis, worker size increases with the number of workers in the colony (Smith & Tschinkel, 2006; Tschinkel, 1987, 1993, 1998). This relationship is at least in part due to increasing resource availability to developing individuals in larger colonies - increased nutrient availability does increase worker production and worker size (Goetsch, 1937; Metzl, Wheeler, & Abouheif, 2018; Passera, 1974; Smith, 2007; Wills et al., 2015). Increased protein/nitrogen content has also been associated with the production of larger workers (majors/soldiers) and sexual castes (Bono & Heithaus, 2002; Goetsch, 1937; Hunt & Nalepa, 1994; Passera, 1974; Schmidt, Hunt, & Smith, 2012; Smith, Anderson, Tillberg, Gadau, & Suarez, 2008; Smith & Suarez, 2010; Smith, 1942). However, previous studies have been largely observational or used natural foods (e.g., increased insect prey) that confound protein/nitrogen content with many other nutrient differences. Furthermore, many previous studies focus on short-term colony dynamics, documenting adult worker survival rather than longer-term growth (e.g., Dussutour & Simpson, 2009, 2012; Wills et al., 2015).
The nutritional basis of individual and superorganism size is a major unresolved issue in the understanding of social insect growth and development. In the present study, we used ants as a model superorganism and employed a nutritional geometry framework (Simpson & Raubenheimer, 1995) that allowed for the manipulation of both the ratio and amounts of protein and carbohydrate available to colonies. We hypothesized that worker size and colony size would both increase with increasing availability of protein in the diet – that is, worker number and size are both protein-limited. This hypothesis is consistent with sociometric observations. An important alternative hypothesis is that worker and colony development are limited by different nutrients (i.e., worker and colony optima differ). For example, worker size may indeed be protein-limited, but colonies may be more carbohydrate limited due to the metabolic demands of the standing crop of adult workers. This alternative hypothesis is premised on there being a nutritionally mediated life-history trade-off between growth (new tissue) and maintenance (current tissue, Fig. 1).
A division of labour inherent in social insect colonies is between growing individuals (larvae) and those that maintain the colony (workers). To further explore possible trade-offs between growth and maintenance, we examined the flow of nutrients through these two types of individual, larva and worker, using a stable isotope tracer experiment. We hypothesized that protein would be more rapidly and completely transferred to larvae while carbohydrates would persist longer in workers.
Lastly, we examined natural (field) colony preference for macronutrient ratios to test whether preference matched the nutrient ratios that maximized growth in the lab. We hypothesized that colonies would prefer what maximized lab growth. However, an alternative hypothesis is that worker preference is not aligned with colony growth (under lab conditions), but instead is more tuned to the adult worker optimum of high carbohydrates – this may be selectively advantageous by favouring colony maintenance over growth. Furthermore, workers are the collectors of the food and preferences based on their own physiology and current state build from the ancestral condition of solitary insects. Other alternative hypotheses include that colonies respond based on experience or environmental variation, among other factors.