Benjamin de Bivort

Benjamin de Bivort

Professor of Organismic and Evolutionary Biology
Department of Organismic and Evolutionary Biology
Center for Brain Science, Faculty of Arts & Sciences
BenDeBivortPhoto

We study the neural circuit and molecular basis of individual differences in behavior using custom high-throughput instrumentation, with an eye to its ultimate cause in evolution.

The animal kingdom contains staggering morphological diversity, but even greater variety is manifest in animal behavior. All animals display species-specific ecological behaviors (such as preferentially interacting with members of the same species), and behavior alone can distinguish species that are otherwise morphologically identical - e.g. two cricket species in the genus Gryllus that inhabit the same range and ecological niche but are distinguished only by distinct mating songs.

Moreover, evolution and behavior exert reciprocal influences on each other - while evolution can diversify behavior, behavior can constrain the evolution of species. For example, in sympatric speciation, behavior provides the reproductive barrier between sub-populations whose hybrid offspring have reduced fitness. In another classic example, two species of periodic cicada with overlapping range emerge to mate after different prime-numbered year intervals, a behavioral strategy that reduces the evolutionary pressures associated with multiple swarms emerging in the same year and periodically abundant predators.

We have previously worked on projects in several other fields of biology, including: cell biology, microbiology, metabolism, systems biology, and arthropod systematics.

What genetic changes underly the evolved differences in behavior between related strains and species? A number of classic fly behavior papers (Dobzhansky et al., 1974; Levene et al., 1976) showed that natural populations of Drosophila pseudoobscura could be artificially selected for positive and negative phototaxis over a small number of generations.

This genetic plasticity likely has mediated the strain and species-level differences in phototaxis we have observed using a number of assays. We are currently using quantitative trait localization to identify the genetic loci responsible for the behavioral differences between several pairs of Drosophila species and strains.

What underlies the behavioral differences between genetically identical individuals? One can consider behavioral diversity at a number of scales. Arthropods may on average behave differently than vertebrates - sister species differ behaviorally either by drift or selection for mating isolation - the behavior of strains within a species may vary due to differential allelic frequencies or phenotypic plasticity - and genetically identical siblings display non-heritable behavioral differences. In humans, this is called personality.

We have found that even in isogenized stocks of fruit flies, individuals display idiosyncratic behaviors in a number of paradigms. In all strains tested, whether they prefer on average to run toward light or darkness when startled, some individuals are more photo-positive or photo-negative than their clonal siblings. While not showing a species-level bias, individual flies prefer to choose either left or right turns in branching mazes. These idiosyncrasies are not heritable, but last the lifetime of the flies, and therefore constitute a form of personality.

A number of arthropod parasites modify their hosts' geotactic behavior. In the case of fungal parasite Entomophthora muscae, infection very quickly induces the flies to climb to the top of vegetation. There it consumes the animal and sporulates, capitalizing on its exposed position to achieve greater spore dispersal. Previous work has shown that Entomophthora can be cultured using traditional microbiological techniques and infect Drosophila melanogaster under laboratory conditions

We are working to establish this relationship as a model system of parasitic behavioral control. We will then investigate whether geotaxis Polarity Control Neurons (previously identified by our group) are in the same circuitry targeted by Entomophthora.

Is there a basic behavioral vocabulary? Careful characterization of both individual and species-level behavioral differences requires high-throughput quantitative characterizations of behavior. We have developed a number of instruments and methodologies to do this, including an imaging rig that allows us to automatically record the position of a single fly's legs as it runs on a floating ball. From this data we can automatically characterize its behavior using both supervised and unsupervised classification algorithms. The latter approach gives us the list of motor primitives being implemented by the animal, and a target list for identifying the central pattern generators responsible for behavior at the lowest level.

Selected Publications:

Ring attractor dynamics emerge from a spiking model of the entire protocerebral bridge.  Kakaria K, de Bivort B. Frontiers in Behavioral Neuroscience. 1, 8, doi: 10.3389/fnbeh.2017.00008, 2017.

Systematic exploration of unsupervised methods for mapping behavior.  Todd J, Kain J, de Bivort B.  Physical Biology. 14, 015002, doi: 10.1088/1478-3975/14/1/015002, 2017.

Recovery of locomotion after injury in Drosophila depends on proprioception.  Isakov A, Buchanan S, Sullivan B, Ramachandran A, Chapman J, Lu E, Mahadevan L, de Bivort B.  Journal of Experimental Biology. 219, 1760-1771, doi: 10.1242/jeb.133652, 2016.

Evidence for selective attention in the insect brain.  de Bivort B, van Swinderen B.  Current Opinion in Insect Science. 15, 9-15, doi:10.1016/j.cois.2016.02.007, 2016.

Social context modulates idiosyncrasy of behaviour in the gregarious cockroach Blaberus discoidalis.  Crall J, Souffrant A, Akandwanaho D, Hescock S, Callan S, Coronado M, Baldwin M, de Bivort B.  Animal Behaviour. 111, 297-305, doi:10.1016/j.anbehav.2015.10.032, 2016.

Contact Information

Northwest Building, Room 235.30 (Office)
Room 248 (Lab)
52 Oxford Street, Cambridge, MA 02138

Faculty Alphabetical