Persistent loss of animal diversity on a rocky shore over nine decades and across multiple investigators.

Long-term ecological monitoring inevitably requires a 'passing of the quadrat' from one investigator to another. Here we present the challenge and opportunity of inferring biodiversity change over time and across investigators using a rocky intertidal case study. An intertidal transect was established and first surveyed in 1931-1933, then resurveyed in detail in 1993. After 1993, the transect was surveyed 16 more times, with surveys nested within four distinct investigator eras. We addressed two goals with our dataset. First, we used a causal framework to detect temporal change in biodiversity (species richness, Hill-Shannon and Hill-Simpson diversity, and species evenness) and then attribute biodiversity change to seawater temperature. Second, we tested the hypothesis that population change in abundance was associated with the geographic range of species. Over thirty years (1993-2023), species richness showed no linear trend but was associated with investigator era; we hypothesize that sampling effort contributed to this effect. In contrast, we observed weak linear declines in the diversity of common taxa (Hill-Shannon), dominant taxa (Hill-Simpson), and evenness over the recent thirty years, but no evidence of investigator effects. We estimated a weak negative effect of maximum sea surface temperature on Hill-Shannon diversity (-0.25 common taxa °C