An automatic feeder splitting the same daily ration into 36 meals instead of two shortened female turquoise killifish lifespan by nearly a third, with no effect on males
An automatic feeder splitting the same daily ration into 36 meals instead of two shortened female turquoise killifish lifespan by nearly a third, with no effect on males
On September 14, Frontiers in Aging published a study from the Medical University of Vienna: the team built an inexpensive automatic feeder for aquariums housing the turquoise killifish, a fish that lives only a few months and serves as a laboratory model of aging, and compared feeding regimens ranging from two to 36 meals per day using three types of dry food and live bloodworm. The total daily amount of food was kept the same for automatic and manual feeding; only the frequency of delivery changed.
The species itself, Nothobranchius furzeri, inhabits seasonal pools in Zimbabwe and Mozambique that dry up for several months each year: its entire life, from embryo to old age, fits within four to six months, which is why laboratories have used this fish for over a decade as the fastest vertebrate model of aging. Automatic feeders for killifish already existed, but the authors considered them expensive, difficult to assemble, or imprecise in dosing, so the Vienna team built their own: a motorized arm with interchangeable food tubes driven by a stepper motor and an Arduino microcontroller, capable of dispensing food with milligram accuracy at a cost of roughly 500 euros per hundred aquariums, cheap enough to feed hundreds of fish simultaneously and support more experiments.
Frequent automatic feeding accelerated growth in fish of both sexes more than the same amount of food delivered manually twice a day; food composition had almost no effect on growth rate when frequency was held constant.
Feeding frequency affected lifespan differently depending on sex, although both sexes started out the same way: in fish of both sexes, frequent automatic feeding initially raised mortality during the first weeks of life. In males, this spike leveled off later, and the median lifespan remained unchanged (20 to 22 weeks), even though automatically fed males grew larger than controls. Growth accelerated, but lifespan did not shorten, contrary to a consistent principle of gerontology: more calories and faster growth are associated with shorter life. In mice this has been confirmed repeatedly, and in this same fish species it recently held in the most literal sense: an edited vgll3 gene accelerated growth and maturation in males while cutting lifespan by 15%. In females, the mortality spike around week seven only intensified instead of resolving: frequent automatic feeding reduced median lifespan from 21.9 to 15.1 weeks (nearly a third), with the strongest effect on the high-fat feeds Biomar and Otohime.
Frequent feeding also altered reproduction. With automatic feeding on the high-fat Biomar diet, females laid more eggs as early as week nine, but embryo survival dropped sharply and the reproductive period compressed. Bloodworm produced the best reproductive outcome, while the leaner JBL feed yielded the fewest eggs but with almost no embryo losses. The authors suggest that the shortened lifespan of females resulted from early reproductive burden, but they note explicitly that males were housed individually while females were kept in groups of six to eight with one male for breeding: group housing may also have contributed to the sex difference.
The fish strain and husbandry protocol were adopted by the Vienna team from the laboratories of Valenzano and Cellerino, who established this species as an aging model over a decade ago. The study shows that the method of food delivery (until now a technical detail of fish husbandry) can create or conceal lifespan differences that could easily be mistaken for an effect of the intervention being tested.