Main text
INTRODUCTION
Conservation planning usually combines general principles with site-specific rules (Groom et al . 2005). In our review (Arroyo-Rodríguez et al . 2020), we combined empirically-supported general principles to describe biodiversity-friendly scenarios in human-dominated landscapes. One such principle is based on extinction thresholds: as most species require 10–30% forest cover in a landscape for population persistence (Andrén 1994; Swift & Hannon 2010), we suggest that a conservative general guideline would be to maintain ≥40% forest cover in landscapes to ensure persistence ofmost forest species (“40% criterion” hereafter).
Banks-Leite et al . argued that the 40% criterion lacks evidence and is problematic, and advocated for using regionally-defined thresholds to guide conservation and restoration. As discussed below, we found their criticisms unfounded, and the idea of basing conservation actions on unique, regionally-defined thresholds impractical and potentially harmful.
VALUE OF THE 40% CRITERION
We are in a state of planetary emergency, with forests increasingly lost and degraded by human activities. Therefore, identifying general conservation guidelines such as the 40% criterion has never been so urgent and valuable. Banks-Leite et al . argue that applying a “fixed ” threshold to “any given landscape ” is “unhelpful and potentially dangerous ” because extinction thresholds vary among species and regions. These claims are misleading, as we specify in our paper that the 40% criterion is not fixed, and should be adapted to some contexts, such as in tropical regions with low matrix quality (p. 1410 and Fig. 1b in Arroyo-Rodríguez et al . 2020). However, the 40% criterion is adequate for most species in most landscapes.
Banks-Leite et al . also raised concerns regarding the lack of social and economic considerations in our guidelines. However, the stated aim of our review was to illustrate an optimal human-modified landscape for biodiversity conservation . We agree that alternative scenarios may optimize other outcomes. Ideally, a multicriteria planning process would evaluate tradeoffs among different scenarios (e.g. Neugarten et al . 2016; Vollmer et al . 2016), and we see great potential for future research to inform such planning.
SUPPORT FOR THE 40% CRITERION
We disagree that the 40% criterion lacks evidence. We cited two global reviews of dozens of studies indicating that most species have extinction thresholds at 30% habitat amount (Andrén 1994; Swift & Hannon 2010). By suggesting 40% (and not 30%), we attempt to be more conservative, and compensate for variation and uncertainties associated with the estimation of habitat thresholds (Rompré et al . 2010). Interestingly, two of the studies we ‘ignored ’ also support a 40% criterion: Macchi et al . (2019) demonstrate that most (71%) forest birds in the Chaco region have extinction thresholds at >38% woody cover, and Melo et al . (2018) found that extinction thresholds average 34% for birds at tropical latitudes. Although Banks-Leite et al . (2014) found that 30% forest cover would safeguard Brazilian Atlantic forest vertebrates, other studies in the same biome document higher thresholds for woody plants (40%, Rigueira et al . 2013) and forest-specialist birds (46%, Morante-Filho et al . 2015). Therefore, a 40% general target is more consistent with the evidence than the 30% preferred by Banks-Leiteet al .
FEASIBILITY OF THE 40% CRITERION
Banks-Leite et al . suggest that economic constraints in regions requiring massive restoration efforts make the 40% criterion unfeasible. We disagree. In 50% of Earth’s forested biomes we have already reached the criterion or have sufficient unaltered forest to meet the criterion, and 27% more biomes have sufficient forest cover to reach the 40% target with restoration (Dinerstein et al . 2017). In the 23% of biomes in which forest cover is < 20% (Dinerstein et al . 2017), we agree that a lower threshold is likely a more feasible target.
Banks-Leite et al . also argue that the 40% criterion creates confusion among policy makers and undermines public trust in science. This speculation is baseless, as the scientific discussion about extinction thresholds is not under public dispute. We rather suggest the opposite: scientists, as members of society, have a responsibility to inform policy makers and the general public on appropriate use of science in addressing societal issues and concerns (Lees et al . 2020).
ON THE USE OF REGIONALLY-DEFINED THRESHOLDS
Banks-Leite et al . advocate for regionally-defined thresholds, because we have the “technical capacity ” to determine optimal regional forest cover scenarios. We found this suggestion unfeasible and potentially harmful for conservation, as it would require information on extinction thresholds of a wide variety of taxonomic groups, each of which will likely have a different threshold. As this information is absent in most regions, we simply do not have the time and resources to identify taxon- and region-specific extinction thresholds across the globe. We agree that, when available, regionally-defined thresholds might be more appealing to decision-makers, but if we insist on using region-specific thresholds everywhere, then deforestation and lack of restoration will continue, and species will go extinct while we spend decades collecting data.