A canopy heat-load threshold for fruiting-site retention and lint yield in long-staple cotton, displaced by trehalose-pathway perturbation

Heat at flowering cuts lint yield through shedding of squares and young bolls, usually reported as non-transferable abscission percentages. We asked whether fruiting-site survival is a continuous function of canopy heat load, and what displacing it is worth. Sowing date and drip-irrigation interval gave eight environments over two seasons at one southern Xinjiang site, spanning canopy heat loads (HL, hours with canopy temperature ≥ 32 °C in ten days from anthesis) of 17.9–70.9 h. Two long-staple cultivars ( Gossypium barbadense L.) received water, melatonin, trehalose or validamycin A, alongside a source–sink manipulation. Retention of tagged first-position sites declined logistically with HL, with half-retention heat loads of 54.0 h (cluster-bootstrap 95% CI 51.2–57.1) and 70.2 h (66.4–75.1) in the sensitive and tolerant cultivars; leave-one-environment-out error was 0.035 retention units. Lint yield followed boll number, falling 53% in the water-sprayed sensitive cultivar. The benefit of validamycin A grew continuously with HL: at 70.9 h retention rose from 0.109 to 0.293 in the sensitive cultivar and from 0.274 to 0.409 in the tolerant one, whose treated thresholds are extrapolated beyond 70.9 h. Averaged over cultivars in the three hottest environments, it raised retention by 55% and lint yield by 19%, with fibre properties unchanged. Leaf photosynthesis fell only 17%, whereas boll cell-wall invertase activity and sucrose-to-hexose ratio tracked retention within environments, consistent with a sink-side carbohydrate limitation. One site and two seasons establish the form of the response, not its transfer between regions; recalibrated locally, the threshold can inform cultivar, irrigation and spray decisions.