The fencing paradigm in woodland conservation: consequences for recruitment of a semi‐arid tree

Abstract

Question: How does long-term fencing against large domestic herbivores affect regeneration of the dominant tree, Prosopis flexuosa, and hence the structure of semi-arid woodlands? Location: Woodlands in the Central Monte Desert biome of Argentina, N˜ acun˜ a´n Man and the Biosphere Reserve area (34° 20′ S, 67° 58′ W) and surrounding cattle-grazed ranches. Methods: We compared seedling emergence and survival, the spatial distribution of seedlings and saplings, and the population stage-based structure of P. flexuosa between paired sites inside and outside the Reserve of N˜ acun˜ a´n (Argentina), which has been fenced to exclude domestic cattle for ca. 40 yr. Results: Reserve sites had lower tree recruitment and seedling emergence, in spite of having greater seed production and seedling survival. Outside the reserve, survival was higher for seedlings in high-density clumps than for isolated seedlings. Seedling clumps occurred mostly near adult individuals, where cattle dung was abundant, suggesting an effect of cattle on seed dispersal. The balance between the effects of cattle exclusion on seedling emergence and on seedling survival was reflected in the stage structure of P. flexuosa woodlands, as populations at fenced sites were dominated by adult individuals, whereas those at cattle-grazed areas presented greater proportions of seedlings and saplings. Conclusion: Fencing is a common practice used worldwide to exclude conservation areas from human disturbance. For example, it is assumed that disturbed woodlands may recover after fencing, thanks to increased tree recruitment after the exclusion of large herbivores. However, the actual effectiveness of fencing as a tool for forest conservation in overgrazed environments could be controversial because sequential effects of herbivores during the tree regeneration cycle may hamper predictions of the overall balance of the recruitment process. Our results suggest that dispersal by cattle influences the regeneration of P. flexuosa woodlands, and that the positive effects of cattle on seedling emergence compensate for the negative effects on seed production and seedling and sapling survival. Future management of P. flexuosa woodlands in the Central Monte Desert should consider that tree recruitment is closely related to land use, and that cattle exclusion does not necessarily guarantee woodland recovery in the long term.