Cooperation among microbes tends to be less sophisticated than that seen between multicellular organisms, especially between animals, much less that seen among the cells making up those multicellular organisms' bodies. Cooperation nonetheless does occur among microorganisms, plus can be recognized as a forerunner, either homologously or analogously, to the cooperation seen, for example, among you and me. That is, the cooperation we familiarly witness around us in part either (1) is a direct evolutionary descendant of cooperation as seen among microorganisms, i.e., a homology, (2) is a copy of microbial cooperative strategies, i.e., is an analogy, or (3) instead could be a combination of both. In this section I consider social interactions between microbes as well as related issues of communication among these organisms.
Communication allows coordination of activities including in terms of the exclusion of defectors from one's or from one's population's midst. Communication, however, is not necessarily a prerequisite for cooperation to evolve since, as noted, simply doing nothing (i.e., not becoming planktonic) can be sufficient to allow cooperators to interact preferentially with cooperators. Nonetheless, in considering real-world examples of microbial cooperation, communication as a means of social interaction is a good place to begin, if only because cooperative acts themselves often can be viewed as forms of communication.
The concept of communication is broader than what most of us are used to, i.e., that which we accomplish using phones, Skype, instant messaging, text messaging, Facebook, email, unamplified voice, or, occasionally, paper. Indeed, communication need not be two-way, in the sense of two or more parties both emitting and receiving signals. Instead, communication can consist of only one party emitting a signal and others receiving that signal. By definition that concept is what I hope will occur as I write these words and, indeed, what (ideally) has occurred now that you have read them. More broadly, communication can be viewed not simply in terms of behaviors but instead in terms of the fitness consequences of social interactions, hence our consideration of signaling as part of our initial discussion of microbial cooperation.
Diggle et al. (2007) , in their review of the evolutionary biology of bacterial quorum sensing, differentiate signals into three categories which they call signals, coercion, and cues (I italicize these terms to distinguish them from less-strict usage). With signaling, both sender and receiver of a signal benefit from the signal. This can be viewed as equivalent to a cooperative action on the part of the signaler, though not an altruistic action since the signaler benefits too. Quorum sensing itself may be described as signals in this regard, i.e., a molecular signal released by bacteria that provides an indication, through group action, of bacterial density (the higher the signal density within an environment then the higher the density of signal producers). Calculation of bacterial density can then allow modification of behaviors to behaviors that are better suited to higher-density circumstances. That is, quorum sensing can be viewed as a cooperative interaction among bacteria. Alternatively, there is no reason that a similar signal-reception program couldn't be employed by a single bacterium found within a sufficiently small volume (Redfield, 2002) , though to what extent such single-organism density determination actually occurs, in nature, is an open question.
Perhaps bridging these distinct interpretations is the hypothesis of Hense et al. (2007) who suggest that quorum sensing may be particularly useful to bacteria as it occurs within bacterial microcolonies. In this case high densities of signal molecule can accumulate within what essentially are high-density and in many cases also clonal bacterial populations. That is, quorum sensing as a cooperative action could very well evolve within such a situation since quorum sensing cheaters may to a significant degree be excluded from such populations at least over short or intermediate time frames.
A signal may benefit the sender but not the receiver, and indeed might even be costly to the receiver. This coercion definitely is not cooperative, and in fact may be described as virulent. The release of antibiotics by one microorganism that can negatively impact a second microorganism, for example, can be viewed as a coercive behavior. More consistent with the idea of communication, however, might be the release of noxious chemicals that have the effect of stimulating negative chemotaxis or other behaviors that result in recipients moving away. The recipients still lose by being forced to move from what otherwise might be a beneficial environment (e.g., resource rich) but clearly are receiving a signal, acting on that signal behaviorally, and otherwise finding the presence of that signal to be costly.
A cue benefits the recipient but not the sender of the signal. The signal employed by a parasite or predator for recognizing and otherwise gaining access to a host or prey organism may be described as a cue. More benignly, to the signaler, the release of metabolic waste may be viewed as potentially cuing other organisms within the vicinity, which may then respond in a manner that is adaptive. If nothing else, the cue is supplying information to a second organism as to the current state of that second organism's environment.
There exists an interesting social interaction among certain viruses, one that is related to quorum sensing, and which, depending upon perspectives, may be viewed as a cue, coercion, or signal. In either case the signaler is a virus-infected cell while the recipient is also a virus-infected cell, and the signal "molecules" are the virus particles themselves. The first cell releases upon lysis one or more viruses which then adsorb (attach) to the second virus-infected cell. The attachment of a virus to an already infected cell, a process that can be called secondary adsorption, signals this second cell that the ratio of viruses to cells in its immediate environment likely is greater than 1. This relatively high ratio, in turn, is suggestive that the density of virus uninfected cells is in decline (Abedon, 1990) —as equivalent to the impact on prey survival of having large numbers of predators in the vicinity. This is helpful to the second virus-infected cell because, in some cases, the infecting virus (the primary virus) can respond by altering its life history characteristics to emphasize retention of its current resource (the infected cell), but this change in behavior is done at a cost of rapid access to additional resource, that is, to other cells. Thus, a predator with many prey to choose from may be able to get away with consuming only the "good parts" of captured prey before moving on to the next prey individual whereas a predator with few prey to choose from may be well advised to take the time to consume even less desirable portions of prey individuals, e.g., just as bears preferentially consume the eggs or brains of salmon when salmon are plentiful but more of the fish when these prey are not plentiful (Gende et al., 2001) .
The initially cell-infecting virus, the primary virus, is unable to maximize its exploitation of the cell it is infecting (the equivalent of virus replication) while at the same time maximizing its exploitation of other cells found in its vicinity (the equivalent of virus dissemination)—just as in the previous section we considered conflicts between bacterial vegetative growth and cell dissemination or survival. The virus-density information that comes with secondary adsorption thus can provide a benefit to the receiving infection, but is costly to the sender due to a failure of secondarily adsorbing viruses to successfully infect (Abedon, 1990; 1994) . This process, in its cost to the sender, thus portrays what Diggle et al. (2007) would describe as a cue. Alternatively, the secondary virus can provide a signal that coerces the primary virus to lyse, a yet additional phenomenon in this complicated mechanism which thereby protects clone mates of the secondary virus from future exposure to that infected cell (Abedon, 2009a) . The same interaction even might be described as a signal to the extent that lysis benefits both parties, protecting the first and effecting progeny dissemination for the second. Secondary viruses, though both unconscious and involuntary in their actions, may be described as displaying cooperative behaviors in the sense that they serve to protect clone mates from also secondary adsorbing.
Diggle et al. (2007) also provide a summary of social traits that occur among microorganisms. Their list is informed by their interest in quorum sensing, so while rather extensive though incomplete, it nevertheless serves as a good complement to the bulk of the discussion otherwise provided here, which I summarize below. Also, it is important to realize that the provided examples of microbial social interactions are not necessarily altruistic nor even necessarily cooperative (i.e., as considered in the previous sections). All of these social interactions, however, can be said to at least be representative of cues if not actual signals.
Table: Communication and other Social Interactions
|Diggle et al. (2007)||Sender-Receiver||West et al. (2006)||Actor-Recipient||Signal||Positive-Positive||Mutual benefits||Positive-Positive||Coercion||Positive-Negative||Selfishness||Positive-Negative||Cue||Negative-Positive||Altruism||Negative-Positive||---||---||Spiteful||Negative-Negative|
West et al. (2006) in turn provide descriptions of social interactions such as altruism and mutual benefits, that is, negative-positive (cues/altruism) versus positive-positive (signals/mutual benefits). Altruism in particular is a cooperative act that is costly to the actor but beneficial to the recipient (that is, it is negative-positive in terms of its impact on actor-recipient) whereas mutual benefits, as that term implies, are mutually beneficial to both actor and recipient (that is, positive-positive).West et al. also describe positive-negative and negative-negative interactions as being selfish and spiteful, respectfully, where in each case the first value refers to the impact on the signaler/actor and the second to the impact on the recipient. Selfishness and coercion, both as positive-negative interactions, also are equivalent.
In their list, Diggle et al. (2007) include such issues as antibiotic production, autolysis, biofilm formation, cooperative foraging, immune modulation, nitrogen fixation, quorum sensing, and swarming behavior all as either representing or involving signals. Antibiotic production can be equivalent to the production of extracellular enzymes, a phenotype that is linked only spatially with the producing organism and which in turn, if involving more than one antibiotic-producing cell, can be similar to cooperative foraging. These mechanisms thus only function for the producer to the extent that released factors cannot diffuse too far beyond the releasing organism prior to having an impact on the local, extracellular environment. They are a cooperative actions in the sense that other organisms in the immediate environment may be able to take advantage of these functions without incurring the expense of effecting them.
See also research arguing that Pseudomonas siderophore production, which are involve in iron acquisition, are a form cooperatively exported molecule (Harrison and Buckling, 2009) and that proximity to siderophores may be important to the fitness of individuals that fail to produce the siderophores (Ross-Gillespie et al., 2009) . Quorum sensing can be similarly viewed as being both beneficial and susceptible to cheating (and in fact can serve as a prerequisite for the secretion of additional, beneficial molecules). Note that all of these extracellular factors are examples of signaling/mutual benefits because both the initiators and the recipients of the communication benefit from them. Note also that players that utilize but don't produce the beneficial extracellular factors can be labeled, as with the siderophore example, as cheaters.
Biofilm formation can involve production of extracellular molecules that are not necessarily enzymes nor even proteins. To the extent that being in a biofilm is beneficial to a bacterium, where benefits stem from more than just not being planktonic, then group living may be described as at least potentially cooperative. Exactly what benefits stem from the existing within a biofilm community, however, are not necessarily as obvious as those benefits that can stem from producing the extracellular factors described in the previous paragraph. Autolysis, which can contribute to biofilm development – essential a form of cellular suicide that is coupled with cellular self-dismemberment (Skulachev, 1999) – similarly is not as obviously beneficial, in this case to the cells that are autolysing. Clearly, though, the recipients of nutrients that are released upon autolysis must have at least some potential to benefit. Therefore autolysis might be seen as a possibly altruistic behavior, with the released nutrients a cue. Autolysis also is a behavior that is preferentially helpful to clone mates if it occurs within the context of a clonal microcolony or biofilm. Cooperative siderophore production, too, has been explored within a biofilm context (Harrison and Buckling, 2009) . In the following section I consider additional examples of potential microbial cooperation.