“…Studies that rely on natural history observations, by contrast, can span much longer periods and therefore provide data uniquely suited to understanding natural biotic patterns, as well as their perturbations, at relevant timescales. Poignant examples include the multi‐year coral reef community structure dataset collected from the Buck Island Reef National Monument (Bythell et al., 2000 ); several multi‐decade‐spanning datasets being collected at the Algonquin Wildlife Research Station, including body size, nesting behavior, and nest temperature of turtles (Connoy et al., 2020 ), reproductive performance and survival of Canada jays ( Perisoreus canadensis ; Derbyshire et al., 2015 ), auditory monitoring of white‐throated sparrows ( Zonotrichia albicollis ; Falls, 1981 ), and population monitoring of small mammals (Fryxell et al., 1998 ); and, in an extreme and famous example, the 30‐year (at time of publishing) Darwin's finches dataset developed by Peter and Rosemary Grant (Grant & Grant, 2002 ). Data collected over such timespans are not only crucial for understanding how natural ecosystems function but are also the only way to test whether predictive models based on experimentation or simulation accurately reflect natural processes (Azaele et al., 2006 ).…”