a new study from UCLA finds both biological brains and AI neural networks developed similar behavioral strategies and neural representations when coordinating their actions, suggesting there are fundamental principles of cooperation that transcend biology and technology.

Some termites form symbiotic relationships with fungus. When harmful fungi invade their carefully cultivated crops, these fungus-farming termites fight back with the precision of skilled gardeners, a new study reveals, smothering them in soil clumps enriched with microbial allies that inhibit fungal growth. Fungus-farming termites, like Odontotermes obesus, maintain a vital symbiotic relationship with the fungus Termitomyces, cultivating it in specialized nutritional substrates called combs that provide both a reliable food source for the termites and an ideal habitat for the fungus. However, these nutrient-rich combs also attract invasive fungal weeds, particularly the fast-growing Pseudoxylaria, which can quickly overtake the crop if left unchecked. While Pseudoxylaria is typically suppressed in healthy combs under termite care, it rapidly spreads when termites are removed, suggesting a critical role of termite activity in maintaining their fungal gardens. While It’s thought that termites use microbial agents to manage these fungal weeds, while sparing their cultivated crop, the precise behavioral mechanism by which they achieve such selective control remains unknown. Through experiments exposing O. obesus to varying severities of Pseudoxylaria outbreaks, Aanchal Panchal and colleagues found that the termites employ a flexible set of behaviors to suppress weeds, adjusting their tactics depending on the severity of the invasion.

When faced with small infections, termites actively remove Pseudoxylaria from contaminated comb and bury it under soil clumps (boluses), which effectively isolates the harmful fungus in an oxygen-deficient soil environment, suppressing further growth. In the case of severe outbreaks, termites fully isolate infected portions from healthy combs and, if necessary, smother entire sections in soil boluses to contain the threat. Notably, the authors found that the soil boluses the insects use are not just barriers – they contain a diverse community of microbes, including termite-derived bacteria with fungistatic properties. Termites deploy these fungistatic boluses only when weeds threaten their gardens, not on healthy fungal combs. According to Panchal et al., this indicates that O. obesus has evolved a highly targeted defense strategy, enlisting microbial allies to selectively combat harmful fungi while sparing their beneficial crop. “The findings of Panchal et al. elucidate how microbial symbionts can be used as part of a multifaceted pest management strategy,” write Aryel Goes and Rachelle Adams in a related Perspective. “Efforts to understand the molecules involved, and their relationship to host fitness, may reveal beneficial microbes that lead to natural product discovery for medicine, agriculture, and bioremediation.”

A global experiment looking at how birds respond to 15,000 paper “moths” reveals that no color-changing strategy to deter predators is universally effective; both camouflage and warning coloration succeed under different ecological conditions, the study shows. Predation is a powerful force shaping evolution, driving the development of two major antipredator color strategies: camouflage, which helps prey to blend into their surroundings to avoid detection, and aposematism, in which prey advertise genuine defenses or, in the case of mimics, deceptive protection, using bright and conspicuous warning colors. Both strategies can be effective under different ecological conditions, yet the environmental factors that favor the development of one adaptation over the other are not well understood, nor have they been evaluated together or across large scales.

To address these questions, Iliana Medina and colleagues conducted a large-scale, global field experiment to test how birds respond to different antipredator color strategies. Medina et al. placed 15,018 paper moth models – some camouflaged brown, some with typical orange-black warning colors, and some with unusual turquoise-black warning patterns – across 21 forests on 6 continents. Each model included a mealworm to attract predators, and consumption was monitored to assess successful predation. According to the findings, no strategy was universally superior; instead, the protective value of each type depended on ecological context. The authors found that camouflage was highly context-dependent, offering early protection under low light or high predator competition. However, camouflage lost effectiveness as predators adapted or cryptic prey became common. In contrast, warning coloration was generally more reliable, though its success was shaped by predator pressure, the frequency and similarity of warning-colored prey in the community, and other ecological factors, especially in lower-latitude environments. Medina et al. show that predator competition is the most influential factor for both strategies and suggest that camouflage may be a less stable defense strategy, one more susceptible to environmental and human-driven changes. This instability could explain why camouflage is gained and lost more frequently than warning coloration over evolutionary time.