Dried-out soil dishes at Caltech selected for more antibiotic-resistant microbes

The apparatus was two sets of soil dishes, some waterlogged and some bone-dry, all of them holding soil microbes and no plants ^[1]. That is a readable comparison: the wet dishes are the control, and the variable being moved is water. What came back from the dry side was microbes evolving greater antibiotic resistance ^[2], with drying itself doing the selecting ^[3]. That is a different pathway from the one that accounts for most resistance, which is human misuse of antibiotics pushing bacteria to evolve genes that ward off the drugs ^[4].

The thing this account does not tell you is how big the effect was, which organisms did it, or which resistance genes they picked up ^[18]. Nor does a soil bacterium shrugging off a drug in a dish mean a patient's infection stops responding to one; the step from selection in soil to a clinical failure runs through gene transfer and human exposure, and none of that is measured here ^[18]. Dianne Newman, the Caltech microbiologist who co-authored the work, put the expectation in geographic rather than numerical terms, saying more resistant bacteria can be expected to be selected for in certain places around the globe ^[5].

The tick figures in the same account are better counted and worse behaved. The 100-fold rise in positive alpha-gal antibody tests between 2013 and 2024 works out to roughly 52 percent a year compounding ^[16], which is steep, though the number is positive tests rather than a rate per test ordered, so testing volume is inside it ^[9]. Nearly half a million people are treated for Lyme disease annually, against more than 8,000 recorded cases of other tick-borne diseases ^[11] - about a 60-fold gap ^[17], and one that mixes a treatment estimate with a case count, so it is not a like-for-like ratio. Tick-borne illness is likely underreported anyway, which means today's true total is unknown ^[14]. The best-quantified projection is the least reassuring for the warming story: a 2022 study found major increases in Lyme only at 3.5 to 5.5 degrees Celsius of warming by 2100, a scenario now considered implausible ^[12], and its authors reported wide uncertainty driven by behavioural variables such as how often people go outdoors and what precautions they take ^[13].

The ranking runs the opposite way from what the framing implies. Ticks carry 90 percent of US vector-borne disease ^[15] and their expansion northward and later into the year is the CDC's stated mechanism ^[8], but the disease burden that follows is loosely pinned. The soil result has a clean comparison and an unreported magnitude. Both remain unquantified as human risks, and this account reports no surveillance system that watches drying soil as a resistance driver ^[18]. Marina Romanello of University College London's Institute of Global Health made the general version of the point, that no individual metric does justice to multiple stressors compounding at once ^[6].

The number worth waiting for is narrow: how often a resistance trait selected in drying soil ends up in an organism that infects people. Until someone measures that, drought belongs on the list of candidate drivers to sample for, not on the list of quantified ones.