Foraging and sex ratio dynamics in C. elegans

Fred Adler, Mirjam Kotar and Sam Webb

August 14,1995

1 Introduction

The nematode Caenorhabditis elegans has been widely used as a study organism in genetics and physiology. Our goal is to build upon this information and begin to think about the ecology of this interesting organism. The dream is to understand the selective forces acting on this organism, to uncover the genetic basis for variability in traits under selection and to analyze how these genetic changes translate through physiology into meaningful phenotypes.

We are working on two major classes of traits, foraging and sex ratio. Foraging includes movement, sensory ability, and ``strategic" components. In addition, important life history decisions depend on the interaction between foraging, reproduction, dispersal and dormancy. In every case, the selective environment is determined by the dynamics and distribution of resources. We thus expect different strains of worms, coming from different environments, to have distinct strategic responses. The patterns of these responses will make testable predictions about their ecological circumstances.

Sex ratio decisions also depend sensitively on ecological factors. The low ratio of males is probably explained by resource distribution and persistence. Differences in dispersal and dormancy between males and hermaphrodites and in male production among strains will produce the key evidence to help understand sex ratio evolution.

The traits we will be studying are quantitative traits, varying continuously within populations due to the action of many genes. These quantitative traits might help us begin to unravel the interactions among genes, the genetic architecture, perhaps through the construction of recombinant inbred lines.

2 Foraging behavior

Foraging theory depends on two components: environment, constraint, and optimization. The environment is the distribution and dynamics of resources. Constraints are physiological (e.g.\ sensory, locomotor, longevity) and cognitive (e.g.\ memory, learning). Given these constraints, one can use models to determine the behaviors that optimize fitness or some convenient surrogate. The hope is that these optimization models will qualitatively (and perhaps quantitatively) predict the responses of organisms to changes in their environment or their physiology. Conversely, strains that behave differently can be predicted to come from different environments.

2.1 Patch departure times

The basic concept of foraging theory is the Marginal Value Theorem, which describes when a rate-maximizing organism should leave a patch as a function of patch quality and expectations about the environment. Our goal is to create a range of patch qualities, and observe how different strains behave. Patches will be distributed by micropipette on 10 cm plates, and incubated for different lengths of time. Ideally, the thickness of the cultures from which the bacteria come could be estimated in advance, otherwise the control will be simply that the plates were poured, seeded and incubated at the same time. To minimize spreading of the bacteria from their original locations, and to guarantee that resources are indeed depleted, we will use chemotaxis plates.

To avoid complications with egg-laying, a single L3 hermaphrodite larva will be placed at the edge of one of these resource dots. We will observe the behavior of this worm, recording the duration and distance of excursions from the dot, and the time when the worm leaves. By so doing, we hope to distinguish between two reasons for shorter residence times: as a strategic departure in response to depletion and as a consequence of exploratory behavior (itself perhaps responsive to resource levels). By photographing and digitizing the tracks left, we should be able to quantify these differences.

The expectation is that strains will differ in their propensity to leave declining resources. Those strains which evolved in a richer environment should ``expect" to find richer resources around them, and should leave more quickly. In addition, we could use various chemical attractants to study the response of worms when they have can estimate environmental quality, however misguidedly. Differences between strains will provide a baseline estimate of the habitat quality these strains experience.

The effects of social factors on foraging behavior can be tested by manipulating the numbers of worms in each patch, in addition to the quality. The various unc mutants could be seeded on patches. It might also be possible to find a combined eat and unc mutant which could neither move nor eat, but which would smell like worms. An alternative control could use worm smell alone (either dauer pheromone or some other extract) in the absence of worms. Optimization theory, here reincarnated as game theory, becomes more complex, but could be used as above to gain insight into worm motivations.

2.2 Patch departure with egg-laying

For real worms, short-term foraging decisions have long-term consequences because eggs are laid in food, and swiftly hatch into hungry larvae. These larvae, furthermore, can become dormant, projecting the consequences of decisions even farther into the future. These decisions could be even more sensitive to social factors than patch departure decisions.

The protocol for these experiments will be identical to the previous ones, except that young adult hermaphrodites will be used, and particular attention will be paid to the spatial and temporal pattern of egg-laying.

In theory, mated and unmated worms should behave differently due to differences in relatedness among offspring and the prospects of male and hermaphrodite offspring. What these differences are is far from clear at the moment. Truly amusing conflicts of interest could arise if relevant genes appear on the X chromosome, because sons cannot have inherited a copy from dad.

2.3 Spatial foraging and area-restricted search

Patch models ignore the spatial complexity faced by foraging organisms. Real spatial foraging can continuously adjust behavior to respond to changing conditions. The baseline for these models is the foraging of organisms in the presence of a spatially homogeneous cue in the absence of food. The track of a single worm on a chemotaxis plate with a controlled level of chemoattractant should give the foraging strategy of adults in the grimmest circumstances (no smell of food) ranging to a virtual gravy train. The paths of worms should, in theory, switch from relatively straight to more convoluted as food smell increases.

Further assays could use real food rather than smells to assess the fine scale behavior of worms. Very small, localized bits of foods in different arrangements could be used to check whether worms engage in area-restricted search. As above, specific strategies depend on expectations about the environment, and should differ among strains in a way parallel to patch departure decisions. Carefully recording tracks could also establish whether worms respond to state (amount of food eaten) or to direct estimates of average food density (experience).

3.4 Worm waves

The behavior of populations is the integrated consequence of the behavior of individuals. Worms placed in a rich but patchy environment spread slowly across empty patches, at a speed determined by patch departure decisions of the different age classes in response to resource depletion and crowding. Several spatial arrangements are possible, including concentric rings, concentric broken rings (in perhaps 6 pieces each), or a regular grid. The first two could be made with carefully chosen and modified paper cups, the last with a ``porcupine". Careful attention would be paid to the age-structure of the spreading population.

Others have observed population waves on depleting plates. One corner of the plate becomes completely depleted, and the worms cluster along the edge of the remainder of the plate, which is occupied by worms and is fairly depleted of food. This wave progresses, encircles the last remaining food, and explodes when the final patch is gone. This sort of thing should be straightforward to model, and could have ecological consequences. In particular, it would be interesting to establish why worms do not advance ahead of the front.

3 Sex ratio dynamics and theory

The sex ratio of hermaphroditic organisms depends on short-term, game theoretic ecological factors (maximizing numbers of grandchildren) and slightly longer-term evolutionary factors (improved genetics). For each factor, ecological circumstances set the stage. In particular, the mixing of genotypes determines whether males are worth making. Differences in mating strategy, particularly the use of plugs by RC301 and other strains, should have important effects on the optimal strategy.

To establish a baseline understand of sex ratio dynamics, we will follow the dynamics of sex ratio in cultures maintained in a range of conditions. First, we will observe sex ratio in fresh plates, seeing whether the reported decline in the fraction of males occurs only after food levels begin to decline and hermaphrodites accelerate their motion beyond the mating speed of males. Depending on these results, we will establish cultures which can spread by controlled dispersal. We will build long trays, approximately 5 cm wide, with nylon separators between square regions. A culture with even sex ratio will be established in one end, and allowed to grow for some time. The separators will be lifted and worms allowed to disperse into the new region. We expect that differences in dispersal propensity between males and hermaphrodites should produce a gradient of the sex ratio. We expect that certain regimes, such as those with short residence durations and long dispersal durations, will maintain a higher male sex ratio.

4 Other experiments

A great deal of ``natural history" about these worms is unknown. The tradeoff between production of males and hermaphrodites could be affected by their different dormancy abilities. Apparently, there has been no consideration of the survivorship of male dauers relative to hermaphrodites.

Lab Home Page