Can slime molds remember?

Unique signals may propagate through the slime mold's tendrils when they reach food

 Seiya Ishibashi / Wikimedia 

Slime mold, Physarum polycephalum, is famous for its seeming intelligence in completing mazes and recreating the map of railways. Sometimes called The Blob, this unicellular network of tubes grows in intricate designs that slowly pulsate and crawl across rotting logs. They eat dead plants and will reconfigure their structures to approach food — enlarging their tendrils toward snacks. Researchers have described this behavior, where a slime mold repeatedly positions near a food source, to reflect formed memory. A pair of scientists, Mirna Kramar and Karen Alim, wanted to understand how a network of tubes can remember. Kramar and Alim's results appeared in PNAS in March.

Previous explanations, such as mechanisms involving epigenetics and altering gene expression, would require at least 30 minutes to establish a non-neuronal memory — too long to fit slime mold learning observations. The researchers made mathematical models to explain how slime mold changes its shape, and they found that food triggers a progressive change in tendril diameter that is slower than the speed of pressure propagation but faster than speed of diffusion. Therefore, perhaps a signal is being transported in the fluids within the tendrils. This signal was proposed to be released when contacting food, increasing tendril diameter while simultaneously shrinking the distant tendrils (because the total fluid volume within all of the tendrils remains constant).  

Although Kramar and Alim did not directly test the nature of this signal, they proposed that it is a chemical that softens the tendril structure, such as adenosine triphosphate (ATP). ATP levels have been measured to be double in slime molds at their migration front than the trailing back. Softening tube walls helps the slime mold grow or spread when migrating. 

Researchers use “memory” more broadly than in colloquial terms, but are slime molds really exhibiting associative memory, or are they just reacting? The mentioned study suggests that slime mold exhibits encoded memory because their shape persists longer than transient reactions. Its corresponding commentary gently initiates a philosophical discussion for distinguishing memory from reactions.