Fuerteventura, one of the eight major Canary Islands, was not the “desert in the ocean” it is today throughout the Early to Middle Holocene.
Scientists (Sánchez-Marco et al., 2026) have recovered the remains of several bird species known to reside at the edges of bodies of water (e.g., lagoons, lakes, rivers) with riparian vegetation and dense forests from a cave in Fuerteventura, the most arid of the Canary Islands. The bones date to ~9000 to 5000 years ago.
This discovery “unexpectedly” reveals the Holocene climate was much warmer (as much as “3 to 7°C”) than present. It was also “much wetter than it is today” a few thousand years ago, and thus regions that are today arid and largely uninhabitable were recently able to host to far more plant and animal species diversity.
The cooler Fuerteventura environment is today covered in sand dunes and classified as an arid desert, as it receives only 100-150 mm of rain annually. The island no longer supports water fowl habitat or any other species dependent on large annual rainfall totals.
“Recent ice core analysis from northern Greenland reveals that the highest Holocene temperatures would occur between 10 and 7 ka BP, 3 to 7 °C warmer than today. This suggests that the animals studied here died in warmer conditions than those prevailing today.”
“It seems likely that there was a lagoon or pond near the cave, around which large areas of riparian vegetation developed. Likewise, wooded areas with undergrowth, where there were even wrynecks, were probably also in the vicinity of the cave. The ornithological record from Cueva del Llano suggests that in the early stages of the Holocene, the dominant climate in the Canary Islands was much wetter than it is today. In Fuerteventura, there were bodies of water with riparian vegetation and more or less dense forest areas with shrubby undergrowth. Higher global temperatures than the present ones may have led to changes in the annual displacements of the Azores High and promoted a more intense rainfall regime, which fostered the maintenance of more diverse habitats and, consequently, a significantly more diverse avian fauna than today. The birds linked to these habitats likely disappeared with climate changes, which led to notably more xeric conditions.”
Thanks.
I came back to this piece after thinking about a quote from an unrelated science video. “Absence of evidence is not evidence of absence” I asked AI to summarize my contention that AGW climate science is presuming just that.
AGW Climate Science claims Absence of Evidence is Evidence of Absence
Thesis
Treating failure to find a fully specified, quantified alternative causal mechanism as proof that none exists is an illegitimate inference that materially inflates confidence in AGW attribution; critical observational and modelling gaps mean many plausible alternatives remain undetected.
Why this is a structural epistemic flaw
Detection‑power requirement — absence only counts when search is adequate.
Multi‑scale nonlinear feedbacks can conceal causation from under‑resolved models or sparse observations.
Convergence risk — shared biases or common assumptions across datasets/models can mimic independent confirmation.
Key empirical gaps that enable the flawed negative inference
Clouds and aerosol–cloud interactions — limited vertical microphysical observations and poor quantification of indirect aerosol effects.
Water‑vapor–convection coupling — sparse oceanic humidity profiles and unresolved convective processes.
Low‑frequency internal variability — instrumental records too short; paleoproxy sparsity limits detection of centennial–millennial modes.
Regional circulation regime shifts — inadequate paleo reconstructions of indices (e.g., Azores High) that can drive persistent regional change.
Biogeophysical land‑surface feedbacks — insufficient, well‑dated vegetation/albedo and groundwater records.
Unmeasured forcings/agents — incomplete dust/aerosol, biological aerosol, and teleconnection records.
Why current model‑based “ruling out” is insufficient
Shared parameterizations and tuning create correlated errors; sub‑grid parameterizations hide process uncertainty; incomplete boundary conditions and low resolution can systematically exclude valid alternatives.
Minimal acceptance criteria any alternative mechanism must meet to supplant AGW attribution
Temporal alignment with proxies; quantitative sufficiency to produce inferred magnitudes; spatial coherence across regions; parsimony (not ad hoc coincidences); replicability in independent data and models.
Critic’s conclusion
Because major observational and modelling detection gaps remain, using the absence of a currently specified alternative as evidence that no alternative exists is epistemically unsound; assertions that “no other science explains warming” are premature when framed as definitive rather than conditional and probabilistic.