Restoration islands are concentrated plantings in strategic locations, created to efficiently use resources to achieve restoration goals. These methods have been used effectively in mesic ecosystems, particularly tropical forests, where the goal of island plantings is often to “nucleate” across a degraded area, providing a seed source for spread outside the planted area. Here, we consider how an island strategy might be used to achieve restoration goals in dryland ecosystems, where limited resources and large areas of degraded land make restoration extremely challenging. In contrast to more productive areas, spread or “nucleation” from restoration islands in drylands may not occur or occur more slowly than required by most management time frames. Despite this, small‐scale, more intensive island plantings may still be useful for achieving short‐term goals, such as weed control, fire management, erosion control, and creation of wildlife habitat. Over the long term, island plantings could serve the same nucleation function as in other ecosystems and serve as repositories for genetic diversity within highly fragmented native systems. Here, we highlight the opportunities for using these high‐intensity, targeted planting methods in dryland ecosystems, provide the guidelines for establishing islands to achieve short‐ and long‐term restoration goals, and identify the areas where additional research is needed to understand the value of restoration islands in dryland ecosystems.

Widespread degradation of natural lands has created an urgent need for restoration. However, the high cost of conventional techniques limits the extent and success of restoration efforts. As a result, practitioners have developed new cost‐effective techniques. Spatially patterned restoration methods, where established clusters of plant species serve as propagule sources across a broad target area, have been proposed as practical restoration techniques. The spatial patterning is expected to reduce initial costs and provide ecological benefits such as increasing habitat heterogeneity. Over the past three decades, multiple spatially patterned restoration methods have emerged around the globe; however, it is unclear whether applications and theoretical foundations have been connected across methods. We conducted a literature review and bibliometric network analyses to (1) examine patterns in focal study systems, cost‐effectiveness, and ecological outcomes for spatially patterned restoration methods and (2) analyze connectivity among the bodies of literature associated with common spatially patterned restoration methods to identify knowledge gaps and synergies. We found the three most commonly studied methods are applied nucleation, slot seeding, and strip seeding. Applied nucleation studies mainly occurred in tropical forests and emphasized plant diversity and seed‐dispersing animal visitation. Slot‐seeding and strip‐seeding studies both primarily occurred in temperate grasslands and emphasized plant establishment and production. Applied nucleation and slot‐seeding approaches had distinct theoretical bases, as evidenced by patterns in reference citation, while strip‐seeding approaches did not draw from a unified body of literature. We discuss the need for full economic analyses and theoretical links between the different methods.

No abstract

The reestablishment and enhancement of plant diversity is typically a priority for restoration practitioners. Since diversity and stability can be affected by the magnitude to which randomness drives community dynamics, modifying randomness (via habitat heterogeneity) could provide utility for vegetation managers. We investigated the value of using strip seeding to manipulate the magnitude to which randomness structures plant communities across a grassland in Davis, California. Five years after restoring portions of a degraded site (0, 33, 50, 66, and 100% of an area) to create patches of seeded and unseeded strips, we assessed the amount of Jaccard dissimilarity across quadrats within strips and estimated the magnitude to which randomness contributed to community assembly (termed the nugget). We found higher nuggets in the 66 and 33% seeding treatment levels compared to the 0, 50, and 100% seeding treatment levels. In the 33 and 66% level of the seeding treatment, we also found that unseeded strips, which are regularly exposed to random events of dispersal from seeded strips, had a higher nugget than seeded strips. This work suggests that strategic seeding techniques that enhance habitat heterogeneity can increase the role of randomness in community dynamics. Strip seeding strategies appear to provide utility as a tool to indirectly enhance diversity across a degraded site.