What is the main unresolved gap in SAR-based extinction rate estimation

Answer

This corpus documents two distinct, still-unresolved limitations of SAR-based extinction-rate estimation, and it does not converge on a single one as definitively "the" main gap. First, extinction-debt / relaxation-rate uncertainty: standard SAR-based estimates assume the system reaches its new equilibrium species number essentially at once after habitat loss, but Halley et al. (2016), in a meta-analysis spanning five taxonomic groups, show that relaxation toward that new equilibrium can take decades to millennia, with half-life increasing with remnant habitat area -- an exponent close to a half for mammals, birds, reptiles and plants when area is indexed by average individuals per species before habitat loss 29816d4c. Because "there is currently no agreement on the rate of relaxation and the factors that it depends on," SAR-derived equilibrium species counts cannot be converted into a validated time-resolved extinction trajectory -- SAR predicts an eventual number of losses but not when they occur 60889779. Second, whether SAR itself is a biased estimator independent of timing: He & Hubbell (2011) argue that the backward SAR method systematically overestimates extinctions by up to 160% relative to actual losses in tropical-forest data, because the endemics-area relationship (EAR) that governs which species actually go extinct is always lower than the SAR used to predict them -- a passive-sampling artifact 70d38dcb. Pereira et al. (2012) counter that this overestimation is not general: whether SAR overestimates depends on the geometry of habitat destruction and the spatial scale considered, and empirical validation of either position is nearly impossible because extinctions unfold over decades to centuries d191013b. He & Hubbell (2013) narrow the mechanism further: the backward SAR equation is only mathematically valid when species are randomly distributed in space, a condition that substituting a standard power-law SAR model does not guarantee and that is rarely met in nature, so the overestimation problem persists specifically when that assumption is violated 7714b871.

Summary

The corpus supports naming two separate, still-open methodological gaps rather than one: (a) the field lacks consensus on relaxation-rate and extinction-debt dynamics needed to time SAR-projected extinctions, and (b) it remains disputed whether SAR itself, and specifically the backward SAR method, systematically overestimates extinction counts due to a sampling artifact tied to species' spatial distribution -- a dispute unresolved partly because it cannot be empirically tested on the multi-decade to multi-century timescales extinctions actually unfold over.

Evidence

• Halley et al., 2016 (Nature Communications) -- meta-analysis across five taxonomic groups; extinction-debt half-life increases with remnant habitat area, exponent close to a half; no consensus on the relaxation rate or its determinants. 29816d4c
• Halley et al., 2016 (same study, additional detail) -- extinction debt can exceed imminent extinctions by orders of magnitude, with relaxation timescales from decades to millennia; SAR predicts the eventual equilibrium loss but not its timing, and no validated method converts SAR equilibrium predictions into time-resolved trajectories. 60889779
• He & Hubbell, 2011 (Nature) -- backward SAR with a power-law model overestimates extinctions by up to 160% in tropical-forest data, because the endemics-area relationship is always lower than the SAR, a passive-sampling artifact. 70d38dcb
• Pereira et al., 2012 (Nature) -- the overestimation He & Hubbell describe is not general; it depends on the geometry of habitat destruction and spatial scale, and empirical validation is nearly impossible given multi-decade to multi-century extinction timescales. d191013b
• He & Hubbell, 2013 (Ecology) -- the backward SAR equation is only valid under random species distribution, a condition rarely met in nature; substituting a power-law SAR model violates the required complementary-area sampling design, so overestimation persists under realistic conditions. 7714b871

Methodology

I grouped Halley et al.'s two submissions as one line of evidence (the extinction-debt/relaxation-rate gap) and the He & Hubbell/Pereira submissions as a second, distinct line of evidence (the SAR-bias/sampling-artifact debate), since they address different mechanisms and neither thread references or resolves the other in this corpus. I did not name a single "main" gap because the verified evidence documents two independently supported, still-open limitations; picking one would discard verified evidence for the other, so I reported and separated both instead.

Limitations

• The corpus does not quantify which of the two gaps -- relaxation timing or estimator bias -- contributes more to projection error in practice, or whether they interact (for example, whether extinction-debt relaxation itself depends on habitat-destruction geometry, the variable central to the He & Hubbell/Pereira dispute).
• The He & Hubbell (2011) overestimation figure comes from tropical-forest data only; the corpus does not establish whether it generalizes to other biomes.
• No verified fact in this corpus reports an empirical test resolving the He & Hubbell vs. Pereira dispute over habitat-destruction geometry, consistent with both sources noting that such validation is close to infeasible on realistic extinction timescales.