Trends in active restoration of tropical dry forest: Methods, metrics, and outcomes

Abstract Active restoration may be necessary to the recovery of degraded tropical dry forests, which are highly threatened but relatively understudied compared to other forest ecosystems. A greater understanding of how different treatments affect various performance metrics and general outcomes is needed. A literature review of active restoration from 2000 to 2019 identified 30 tropical dry forest restoration-related peer-reviewed articles. Grazing (50% of studies), fire (37%), and cultivation (27%) were the most frequent causes of dry forest ecosystem degradation, and invasive non-native species were reported present in the majority of studies. Common site management needs thus included control of competitive vegetation, as well as supplemental irrigation. Outplanted species were most often selected for their functional traits (57%) or local abundance (32%), while only three studies focused on reintroducing or enhancing populations of rare, endangered, or sensitive species. Survival was the most common metric used to measure outplanted seedling performance (88%) and ranged from 13% to 80% (mean 45%). Seedlings planted in mixed assemblages had significantly higher survival rates (median 63%) than those in single species plots (29%), and a moderately positive relationship was observed between survival and the number of species planted. Survival was significantly higher for treatments in which non-native vegetation was continuously controlled throughout the study (median 61%), as well as treatments that received supplemental water for a portion of (63%) or the entire (51%) monitoring period. Survival rates tended to be higher in protected sites (54%) than in sites without any protection status (30%), but this difference was not statistically significant. The cause of seedling mortality, reported in 63% of studies, was most often attributed to desiccation (including soil conditions, timing or lack of precipitation, competition for available water). In general, variation in the format and level of detail reported in the results limited the cross-study analysis that could be performed. Metrics other than survival were often underreported, particularly reproductive measures. The majority of papers also did not report project costs (80% of studies) or community engagement (73%). Results suggest that more diverse metrics and outcomes should be regularly reported in active restoration literature in order to improve best practices and broaden the scope of future research. Remotely-sensed metrics, for example, have the potential to complement field-based evaluations and increase the frequency of future monitoring efforts.