The effective collective: Grouping could ensure animals find their way in a changing environment

Prince­ton Uni­ver­sity researchers report in the jour­nal Sci­ence that col­lec­tive intel­li­gence is vital to cer­tain ani­mals' abil­ity to eval­u­ate and respond to their envi­ron­ment. Con­ducted on fish, the research demon­strated that small groups and indi­vid­u­als become dis­ori­ented in com­plex, chang­ing envi­ron­ments. How­ever, as group size is increased, the fish sud­denly became highly respon­sive to their surroundings.

These results should prompt a close exam­i­na­tion of how endan­gered group or herd ani­mals are pre­served and man­aged, said senior researcher Iain Couzin, a Prince­ton pro­fes­sor of ecol­ogy and evo­lu­tion­ary biol­ogy. If wild ani­mals depend on col­lec­tive intel­li­gence for migra­tion, breed­ing and locat­ing essen­tial resources, they could be imper­iled by any activ­ity that dimin­ishes or divides the group, such as over­hunt­ing and habi­tat loss, he explained.

"Processes that increase group frag­men­ta­tion or reduce pop­u­la­tion den­sity may ini­tially appear to have lit­tle influ­ence, yet a fur­ther reduc­tion in group size may sud­denly and dra­mat­i­cally impact the capac­ity of a species to respond effec­tively to their envi­ron­ment," Couzin said. "If the mech­a­nism we observed is found to be wide­spread, then we need to be aware of tip­ping points that could result in the sud­den col­lapse of migra­tory species."

The work is among the first to exper­i­men­tally explain the extent to which col­lec­tive intel­li­gence improves aware­ness of com­plex envi­ron­ments, the researchers write. Col­lec­tive intel­li­gence is an estab­lished advan­tage of groups, includ­ing humans. As it's under­stood, a group of indi­vid­u­als gain an advan­tage by pool­ing imper­fect esti­mates with those around them, which more or less "aver­ages" sin­gle expe­ri­ences into sur­pris­ingly accu­rate com­mon knowl­edge. For instance, the paper in Sci­ence cites a 1907 study that pre­dicted with near pre­ci­sion the weight of an ox based on the esti­mates of 787 people.

With their work, Couzin and his coau­thors uncov­ered an addi­tional layer to under­stand­ing col­lec­tive intel­li­gence. The con­ven­tional view assumes that indi­vid­ual group mem­bers have some level of knowl­edge albeit incom­plete. Yet the Prince­ton researchers found that in some cases indi­vid­u­als have no abil­ity to esti­mate how a prob­lem needs to be solved, while the group as a whole can find a solu­tion through their social inter­ac­tions. More­over, they found that the more numer­ous the neigh­bors, the richer the indi­vid­ual -- and thus group -- knowl­edge is.

These find­ings cor­re­late with recent research show­ing that col­lec­tive intel­li­gence -- even in humans -- can rely less on the intel­li­gence of each group mem­ber than on the effec­tive­ness of their com­mu­nal inter­ac­tion, Couzin said. In humans, research sug­gests that such coop­er­a­tion would take the form of open and equal com­mu­ni­ca­tion among indi­vid­u­als regard­less of their respec­tive smarts, he said.

The researchers placed fish known as golden shin­ers in exper­i­men­tal tanks in groups as low as one and as high as 256. The tanks fea­tured a mov­ing light field that was bright on the outer edges and tapered into a dark cen­ter. To reflect the chang­ing nature of nat­ural envi­ron­ments, they also incor­po­rated small patches of dark­ness that moved around ran­domly. Pro­lific school­ers and enthu­si­asts of dark­ness, the golden shin­ers would pur­sue the shaded areas as the researchers recorded their move­ment using com­puter vision soft­ware. Although the fish sought the shade regard­less of group size, their capa­bil­ity to do so increased dra­mat­i­cally once groups spanned a large enough area.

The researchers then tracked the motion of indi­vid­ual fish to gauge the role of social influ­ence on their move­ment. They found that indi­vid­u­als adjusted their speed accord­ing to local light level by mov­ing faster in more brightly lit areas, but with­out social influ­ence the fish did not nec­es­sar­ily turn toward the darker regions. Groups, how­ever, read­ily swam to dark areas and were able to track those pre­ferred regions as they moved.

This col­lec­tive sens­ing emerged due to the coher­ent nature of social inter­ac­tions, the authors report. As one side of the group slowed and turned toward the shaded area, the other mem­bers did as well. Also, slow­ing down increased den­sity and resulted in darker regions becom­ing more attrac­tive to these social animals.

Couzin worked with lead authors Andrew Berdahl, a Prince­ton grad­u­ate stu­dent, and post­doc­toral fel­low Colin Tor­ney, both cur­rently in Couzin's lab, as well as with for­mer lab mem­bers Chris­tos Ioan­nou and Jolyon Faria, who are now at the Uni­ver­sity of Bris­tol and the Uni­ver­sity of Oxford, respectively.

This work was sup­ported in part by grants from the National Sci­ence Foun­da­tion, the U.S. Office of Naval Research, the U.S. Army Research Office and the Nat­ural Sci­ences and Engi­neer­ing Research Coun­cil of Canada.