The same question in every geological stage

For every stage of the Phanerozoic, the study compared fossil communities containing reef builders or sediment-mixing bioturbators with communities lacking them. It assessed diversity not only by species count but with the Shannon measure, which includes abundance distribution, and calculated an effect size for each stage. Rather than melting roughly 541 million years into one total, the stage-by-stage design shows when the relationship persists and when it weakens.[1]

Through much of the Phanerozoic, communities with ecosystem engineers carried significantly greater diversity, with the strongest effect at moderate temperatures. The association has a plausible mechanism: reefs create three-dimensional shelter and surface, while sediment mixers create flows of oxygen and nutrients. But because the evidence is a comparative fossil record, the mechanism is not experimentally separated; engineers may raise diversity, while diversity may also mark stable conditions in which engineers persist.[1]

The kind of engineer changes the future

Across roughly the past 250 million years, modern reef builders appear more sensitive to climate stress than bioturbators. The two groups do not construct habitat in the same way: a reef is a complex physical structure sensitive to temperature and chemistry, whereas sediment mixing is a more distributed process across the seabed. Treating ‘ecosystem engineer’ as one resilience class under present warming is therefore misleading. Which engineering function disappears may matter as much as the species count.[1]

The fossil record is a sequence of fragments favouring preservable shells, accessible rocks and well-sampled regions. Even so, the same comparison recurring over hundreds of millions of years frames life as an active builder of physical environments. Organisms construct cavities, surfaces and chemical pathways that later life can use. The association serves as a sign that the history of diversity must be written with environment-building organisms, rather than as a verdict of causality.[1]