Using Two‐Stage Random Sampling and N‐Mixture Models to Estimate the Breeding Population Size of a Threatened Nomadic Species in Morocco

Robust estimates of animal population size are critical for effective conservation planning, yet such studies remain scarce in the North African region, making it difficult to assess the population status of various species. Here, we present the results of a national‐scale survey of the rare and elusive Marbled duck ( Marmaronetta angustirostris ) conducted over 2 years in Morocco. Using a two‐stage random sampling design, we surveyed 187 secondary sampling units of 0.25 km 2 nested within 35 primary sampling units of 100 km 2 , spanning from the Moulouya River to the Saharan wadis. We applied hierarchical binomial N‐mixture models to estimate the breeding population size of the species and identify key breeding habitat features. At the scale of sampled sites that cover 46.75 km 2 of wetlands, our best‐supported model estimated a total of 24 broods, with an average detection probability of 0.71 (95% BCI: 0.49–0.86), highlighting the need to account for imperfect detection in field surveys. When extrapolated to all Moroccan wetlands (748 km 2 ), this translated into a national breeding population size estimate of 386 broods (95% BCI: 283–487). Emergent vegetation was found to be the strongest predictor of brood abundance. Among wetland types, ponds showed positive associations with brood abundance relative to dams, while rivers were associated with lower abundance. These findings represent the first robust national‐scale estimate of Marbled duck breeding population size in Morocco. The analysis highlights emergent vegetation and wetland types as predictors of brood abundance, offering direction for habitat‐focused conservation efforts. The scalable survey and modeling framework used here can be adapted elsewhere to improve monitoring and conservation efforts of vulnerable waterbird species.