A glass eel arriving at a river mouth is sexually undifferentiated. Sex is settled later, during the yellow eel stage, and the outcome is environmentally influenced rather than fixed at fertilisation. Across studies, the most consistent predictor is local density.
High-density sites, typically estuaries and lower reaches where many recruits pile up, skew male. Low-density sites, often far upstream where fewer eels penetrate, skew female. Females also grow larger and stay longer in fresh water before silvering and heading to sea, which fits a pattern where a fish in an uncrowded place invests in size and fecundity.
This makes barriers a subtler problem than they appear. A dam that stops eels reaching the upper watershed does not just reduce numbers; it removes exactly the low-density habitat that produces large females, and so cuts egg production more sharply than a headcount would suggest. Because the whole species is thought to spawn as a single population in the Sargasso Sea, any regional loss is a loss to everyone. The physiological trigger linking density to sex is not established.