We analyse the concept of carrying capacity (CC), from populations to the biosphere, and offer a definition suitable for any level. For communities and ecosystems, the CC evokes density‐dependence assumptions analogous to those of population dynamics. At the biosphere level, human CC is uncertain and dynamic, leading to apprehensive rather than practical conclusions. The term CC is widely used among ecological disciplines but remains vague and elusive. We propose the following definition: the CC is ‘the limit of growth or development of each and all hierarchical levels of biological integration, beginning with the population, and shaped by processes and interdependent relationships between finite resources and the consumers of those resources’. The restrictions of the concept relate to the hierarchical approach. Emergent properties arise at each level, and environmental heterogeneity restrains the measurement and application of the CC. Because the CC entails a myriad of interrelated, ever‐changing biotic and abiotic factors, it must not be assumed constant, if we are to derive more effective and realistic management schemes. At the ecosystem level, stability and resilience are dynamic components of the CC. Historical processes that help shape global biodiversity (e.g. continental drift, glaciations) are likely drivers of large‐scale changes in the earth's CC. Finally, world population growth and consumption of resources by humanity will necessitate modifications to the paradigm of sustainable development, and demand a clear and fundamental understanding of how CC operates across all biological levels.

Ecosystems that change through time impose new scientific challenges for fisheries management advice. We present a case study to illustrate our view on how to face such challenges. The Pink Shrimp fishery in the Southern Gulf of Mexico has collapsed. Annual yields were about 24,000 metric tons during the mid-1950s to early 1970s; currently, they are about 1,200 metric tons. Overfishing was assumed as the main cause, but single-species models failed to provide the advice necessary for recovery. An inverse relationship between stock abundance and temperature was demonstrated, and a decline in recruitment and primary production (since 1970s) was observed. We constructed a trophic model for the ecosystem using Ecopath with Ecosim, incorporating the annual mean anomaly of the Atlantic Multidecadal Oscillation as a climate change index to force primary production. Signals were propagated throughout the food web, and biomasses were simulated for the period of 1956 to 2011. Ecosystem changes were estimated with the highest carrying capacity by the mid-1970s then declining with time; Pink Shrimp follows such a decline. Balanced harvesting was simulated and the “ecosystem reference level” was identified as a maximum harvesting of 40% (catch/biomass ratio) for all resources for a sustainable ecosystem. Conventional single-species management resulted in population crashes. Los ecosistemas que cambian en el tiempo, imponen nuevos retos científicos en cuanto a las recomendaciones para el manejo de pesquerías. Se presenta un caso de estudio para ilustrar nuestra visión acerca de cómo enfrentar dichos retos. La pesquería del camarón rosado en el sur del Golfo de México está colapsada. Desde mediados de la década de 1950 hasta inicios de la de 1970, los rendimientos anuales de esta pesquería fueron de cerca de 24,000 toneladas métricas; actualmente se pescan aproximadamente 1,200 toneladas métricas. En un inicio, se asumió que la sobrepesca era la causa principal, sin embargo las recomendaciones derivadas de los modelos poblacionales no fueron útiles para la recuperación de la pesquería. Se encontró que existe una relación inversa entre la temperatura del mar y la abundancia del stock, y un decremento del reclutamiento y de la productividad primaria (desde la década de 1970). Se construyó un modelo trófico de ecosistemas, mediante Ecopath con Ecosim, forzando la producción primaria del sistema con la anomalía anual promedio de la Oscilación Multidecadal del Atlántico, asumiendo ésta como índice de cambio climático. Las señales del índice se propagaron a través de la red trófica y se simularon las biomasas para el periodo 1956–2011. Se estimaron los cambios en el ecosistema, encontrando los niveles más altos de la capacidad de carga a mediados de la década de 1970 y declinando después de eso; el camarón rosado siguió la misma trayectoria. Se simuló una condición de captura balanceada y se identificó un “nivel de referencia del ecosistema” como la máxima captura del 40% (cociente captura/biomasa) para todos los recursos en un ecosistema sustentable. El manejo basado en enfoques monoespecíficos, resultó en el colapso de la población.