Experimental evolution studies attempt to capture the above-noted time variable and associated genetic evolution. Experiments conducted in the laboratory, or even in situ, however, are often shorter, by quite a bit, than the natural experiments that take place outside of the control of researchers. In addition, culture volumes typically are smaller, and thereby so too are population sizes. The result, at least potentially, is less mutational variation upon which selection can act, though an advantage of microorganisms in evolution studies nonetheless is their small size and therefore potential to exist even in small volumes at high numbers. On the other hand, experimental environment often are much simpler than natural environments, which makes natural selection during these experiments more straightforward but at the same time less varied in its effects. Lastly, the potential for horizontal gene transfer is more limited and/or more contrived in experimental evolution studies, since communities typically are less diverse than is the case in the outside world, often consisting of just a single species or single strain. In natural environments it particularly is long time periods, great heterogeneity, and large volumes/population sizes that conspire to make even extremely rare gene transfer events relatively likely versus relatively unlikely in the laboratory. In this chapter I discuss basic principles of microorganism passage through time within the context of either experimental or applied evolution.

Table: Comparing Experimental with Natural Evolutionary Processes

VariableExperimental evolutionNatural evolutionResult (experiments)
Time Often much shorter (days, weeks, months, occasionally years) While naturally occurring evolution can occur over much shorter time frames, often years, decades, millennia or more are available Less time over which all evolutionary force can act; for instance, less diversity, both genetic and in terms of selection
Experimental volumes Though microorganisms are small in comparison with more macro organisms, even liters of a culture is small in comparison to many natural environments Can be extremely large or, given spatial structure, various combinations of very large and perhaps very small Experiments are limited to emphasizing either selection or drift, but not simultaneously both; lower likelihood of especially rare multiple mutation events; can be compensated for to some degree by studying general mutator strains
Environmental heterogeneity Even when high during experimental evolution, environmental heterogeneity tends to be very low in comparison with natural environments A hallmark of natural environments is their heterogeneity, where substantial variance can occur over very small as well as very large scales, and everything in between The kinds of selection that populations are subject to is much more uniform as so too also are population sizes in experimental evolution studies whereas in natural environments populations can consist of a combination of multiple sub-populations of various sizes, displaying qualitatively and quantitatively different genetic diversities, and be subject to different selective constraints
Community diversity Typically migration opportunities are either not very diverse, are otherwise highly controlled, or are suppressed altogether since the diversity of experimental communities typically is substantially suppressed (e.g., pure-culture technique) Every possible migratory opportunity can and will be available to the point where absence of such opportunities seem noteworthy as unusual circumstances (e.g., as is mostly the case with mitochondria) The genetic diversity upon which selection can act during experiments likely is much poorer than that available in natural "experiments", resulting in greater experimental control but also greatly reduced diversity in possible experimental outcomes

There exist two basic approaches to experimental evolution studies. These may be differentiated in terms of whether organism passage through time occurs within continuous versus more discrete volumes, that is, continuous culture versus, in the case of microorganisms, serial passage or transfer. Continuous culture is exemplified by the chemostat, a device in which organism populations are encouraged to exist at an environment's carrying capacity. With serial passage, by contrast, organisms are encouraged to exist in a state of ongoing if episodic increases in population density. In many ways serial transfer/passage experiments are simpler procedures plus, in comparison with continuous culture, serial transfer is the older technology. In terms of logistic growth, selection is more k (carrying capacity)-influenced for continuous-culture methods versus more r (intrinsic growth rate)-influenced for serially transferred, batch-culture methods. Of course, there also exist more hybrid approaches which are both r and k influenced.

Table: Comparing Serial Passage with Continuous Flow Experiments

VariableSerial passageContinuous passage
Difficulty Can be more work intensive to carry out experiments Can be more work intensive to set up experiments
Dilution characteristics Culture dilution occurs episodically Culture dilution occurs continuously, particularly as seen during chemostat growth; this is less true in other approaches to continuous culture (e.g., turbidostat) but nonetheless with these other approaches culture dilution per episode occurs to a lesser degree than with serial passage
Complexity Allows greater experimental manipulation, including use of solid media or tissue culture, multiple organism physiological states, non-identical passage conditions, and passage through highly complex environments such as animals More mathematically tractable; "In biology there are relatively few accepted mathematical models. The chemostat is one — and, in microbial ecology, perhaps the only one — such model that does seem to have wide acceptance." (Smith and Waltman, 1995)
Competition Organisms, at least in part, compete on the basis of rates of population growth rates, often within environments that are not nutrient limited Organisms compete on the basis of population growth rates within what typically are nutrient-limited environments
Selection diversity Depending on experimental design, organisms also can compete on the basis of their rate of transition to greater nutrient availability (lag phase) and on the basis of their transition to nutrient depletion (stationary phase) Chemostats, particularly those containing only two tropic levels (e.g., heterotrophic organisms and their nutrient supply), typically are much more selectively static
Spatial diversity Formation of wall populations (multiple niches) can be less of a concern since growth chambers can be changed repeatedly Formation of wall populations is a legitimate concern since organisms, e.g., bacteria, can be present in growth chambers over long periods (days, weeks, even months)
r versus K selection During especially exponential growth phase, r selection (based on intrinsic growth rates) particularly without nutrient limitation can be an important component of evolution; however, if stationary phase is allowed to occur, then some K selection can occur as based on growth and survival characteristics under the more limiting conditions observed at higher organism densities Because at steady state nutrients are limiting, populations directly compete for limited nutrients during most or all of the course of experiments, resulting in stronger biases towards K selection than one sees during most batch culture experiments; the result can be greater selection for efficiency of nutrient use rather than solely intrinsic growth rates