Identifying how the evolution of colony ontogeny is shaped by ecology remains one of the major challenges of sociobiology. Colony size and caste composition are two colony traits generally predicted to undergo substantial ontogenetic change, with an extended period of continuous growth followed by stability in the representation of the largest colony members and most specialized castes only at reproductive maturity. Yet these predictions are based on classic theory developed and subsequently tested using data primarily from soil-nesting ants. Here, we address the contrasting hypothesis that growth in the arboreal realm, an environment with substantially different ecological pressures, is modular via the addition of repeated nest units, and is coupled with early stability in caste production and size. We do this using full-colony collections of the turtle ant Cephalotes persimilis, spanning newly founded to reproductively mature colonies. We demonstrate predictable modular growth via an isometric scaling relationship between increases in colony size and the number of nests a colony has expanded into, and a minimum nest number for reproduction. This modular growth is further associated with specialized use of cavities with particular entrance and stem sizes. Additionally, soldiers are produced immediately by newly founded colonies occupying a single nest, and percentage soldier composition scales isometrically (19%) across all colony sizes regardless of reproductive status. This stable caste composition at the colony level is contrasted against highly variable deployment of castes across the nests of each colony. Finally, ant size increases steadily for worker and soldier castes in early colony growth and then stabilizes at reproductive maturity, but this transition happens early in colony ontogeny, at only 200 ants. These findings challenge key theoretical and empirical expectations and contribute new insights into how ecology shapes the adaptive evolution of colony ontogeny.

确定生态学如何影响群体个体发育的进化仍然是社会生物学的主要挑战之一。群体大小和种姓组成是两个群体性状，通常预测会发生实质性的个体遗传变化，持续增长的时间延长，随后最大的群体成员和最专业的种姓仅在生殖成熟时表现稳定。然而，这些预测是基于经典理论开发的，随后主要使用来自土壤筑巢蚂蚁的数据进行测试。在这里，我们提出了相反的假设，即树栖领域的生长是一个具有显着不同生态压力的环境，通过添加重复的巢单元进行模块化，并且与种姓生产和规模的早期稳定性相结合。尖头鱼，跨越新建立到生殖成熟的殖民地。我们通过菌落大小的增加与菌落扩展到的巢穴数量以及繁殖的最小巢穴数量之间的等距缩放关系证明了可预测的模块化增长。这种模块化增长进一步与具有特定入口和茎尺寸的空腔的专门使用相关联。此外，士兵会由占据单个巢穴的新建立的殖民地立即产生，并且无论生殖状态如何，士兵组成的百分比在所有殖民地大小上都是等轴测 (19%)。这种在殖民地一级稳定的种姓构成与跨每个殖民地的巢穴高度可变的种姓部署形成对比。最后，对于工蚁和士兵种姓来说，蚁群的大小在蚁群早期生长时稳步增加，然后在生殖成熟时稳定下来，但这种转变发生在蚁群个体发育的早期，只有 200 只蚂蚁。这些发现挑战了关键的理论和经验预期，并为生态学如何塑造群体个体发育的适应性进化提供了新的见解。