The Choosy Herd

Guanacos and vicuñas live in harems — one male, several females, their young. This social structure is widespread among camelids and often explained through male competition: the strongest males monopolize access to females. But Gonzalez, Abramson, and Laguna (2602.22139) propose a different origin. The harem structure emerges from female choice alone.

Their agent-based model gives females a simple decision rule: evaluate your current group against available alternatives, and switch if switching improves your fitness. Fitness depends on local environmental quality and group size — being in a group provides predator protection, but too large a group means competition for forage. Males are passive in this model. They don't fight. They don't display. They simply exist as group anchors.

The result: harem-like social structures emerge spontaneously. Group sizes distribute in patterns that match field observations. Fitness distributions stabilize. And the population-level properties — group sizes, spatial distributions, fitness variation — are modulated by density, sex ratio, and habitat heterogeneity in non-additive ways. You can't predict the combined effect of two factors from their individual effects.

The non-additivity is worth pausing on. In a simple model where females independently optimize, you'd expect population-level properties to be sums of individual decisions. But females are competing for the same groups, creating frequency-dependent dynamics. When one female joins a group, she changes its attractiveness to other females — both positively (larger groups are safer) and negatively (larger groups have less food per individual). This feedback makes the system nonlinear. The population-level pattern is an emergent property of coupled individual decisions, not an aggregation of independent ones.

What's elegant about this model is what it removes. No male-male competition. No female signaling. No kin selection. No learning. Just local fitness evaluation and switching. And still the harems form. This doesn't mean male competition is irrelevant in real camelids — it clearly is — but it establishes a baseline. The harem structure doesn't require male competition to exist. Female choice is sufficient.

There's a broader principle here about the minimum mechanism needed to explain a pattern. Behavioral ecology has a tendency to invoke complex explanations — sexual selection, costly signaling, strategic behavior — for patterns that might arise from simpler processes. The test should always be: what is the simplest model that generates the pattern? If individual optimization with local information produces harem-like groups, then any additional mechanism (male competition, female signaling, kin preference) needs to explain something beyond the basic pattern. The burden of proof shifts.

The modulation by habitat heterogeneity is the most ecologically interesting finding. Heterogeneous environments — patchy resources, varied terrain — produce different group structures than homogeneous ones, even with the same population and sex ratio. The landscape shapes the social structure not through direct environmental pressure but through the decision landscape it creates for individual females. Good patches attract more females, which increases competition, which pushes some females to worse patches where they have less competition. The spatial distribution of resources generates the social distribution of individuals.

This model is for guanacos, but the mechanism is general. Any system where individuals choose groups based on local assessment of quality and size should produce similar dynamics. The question is whether the mechanism explains real field data better than alternatives, and that requires parameterization against specific camelid populations. The mathematics says the mechanism works. Whether it's the mechanism that works in actual herds remains open.