In a paper in press in the Journal of Theoretical Biology, Randal Bollinger and colleagues suggest that the human vermiform appendix functions as a "safe house" for bacteria:
The observations described above, in conjunction with the survival advantages afforded to bacteria by biofilms(Costerton, 1995; Costerton, 1999; Costerton et al., 1995) and the architecture of the human large bowel, give rise to the idea that the appendix is a compartment well suited for maintaining beneficial or commensal microorganisms, being well positioned to avoid contamination by pathogenic organisms present transiently in the fecal stream. Indeed, the narrow lumen of the appendix as well as its location at the lower end of the cecum are both factors that afford relative protection from the fecal stream as it is propelled by peristalsis. Given the metabolic advantages (Bradshaw et al., 1994; Bradshaw et al., 1997) and other advantages (Costerton, 1995; Costerton, 1999; Costerton et al., 1995) that biofilms are known to afford bacteria, biofilm formation in the appendix is expected to be a relatively effective means of preserving and protecting commensal bacteria. In essence, the structure of the appendix is expected to enhance the protective effect of biofilm formation for commensal bacteria. Effective biofilm formation by commensal bacteria in the appendix is expected to facilitate not only the exclusion of pathogens, but also the adherence of the non-pathogenic commensal organisms within that cavity. Regular shedding and regeneration of biofilms within the appendix would be expected to re-inoculate the large bowel with commensal organisms in the event that the large bowel became infected by a pathogen and was flushed out as a defensive response to that infection (Bollinger et al. 2007:7-8).
So in other words, the appendix supposedly facilitates the use of diarrhea as a pathogen-shedding mechanism, because it would preserve the beneficial gut flora when the body sheds the biofilm coating the rest of the colon.
It's a hypothesis.
So here's a problem: The authors note that the appendix may now be relatively useless because of the lower importance of fecal-borne diseases "in the face of modern medicine and sanitation practices." But before people started living in sedentary settlements, with high population densities, fecal-borne diseases must have been vastly less important than they have been historically, when people regularly have drunk from contaminated water supplies. Throughout human evolution up until the Holocene, diarrheal diseases would not have been absent (food poisoning and incidental contact with fecal-borne disease from other species always being possible), but they must have been vastly less important than they have been recently.
Indeed, maintenance of a reserve supply of commensal bacteria in the event of infection by pathogens may be unnecessary in areas where outbreaks of enteric pathogens do not affect the vast majority of the population at any one time. Certainly this idea is consistent with the well-known observation that appendectomy is without currently discernable long-term side effects in societies with modern medical and sanitation practices (ibid.:9).
If it's adaptive, it must be not to the recent environment of high enteric pathogen load, but the ancestral environment. And not just to meat-eating: the apes have appendices, too. In fact, Fisher (2000) points out that many primates, including non-anthropoids, have similar structures. Others have thickened "biofilm"-like mucosal layers in the cecum that would seem to approximate this microbe-harboring function. Since this includes primarily fruit-eating and leaf-eating species, there is no reason why contact with meats, bacteria from spoiled meats, or enteric bacteria from consumed animals should necessarily be involved. I'm not sure why the current paper (Bollinger et al. 2007) didn't cite any of this comparative work; it certainly seems relevant to their hypothesis.
The appendix looks functional to me, I'm just not sure that it is specifically a "safe harbor" in the face of enteric pathogens. It may just be a "safe harbor" in the face of normal elimination.
I guess that last is an unfortunate turn of phrase...
Bollinger RR, Barbas AS, Bush EL, Lin SS, Parker W. 2007. Biofilms in the large bowel suggest an apparent function of the human vermiform appendix. J Theor Biol (in press) doi:10.1016/j.jtbi.2007.08.032
Fisher RE. 2000. The primate appendix: a reassessment. Anat Rec B 261:228-236. doi:10.1002/1097-0185(20001215)261:6<228::AID-AR1005>3.0.CO;2-O