Just read a research highlight on why bee colonies collapse. Briefly, colonies begin to decline when “the number of functionally impaired bees reaches a critical threshold,” even if no bees have actually been killed. This makes it very hard to find the “cause” of colony collapse. Collapse may be triggered by any of a number of things, and indeed the degradation of the colony may itself be a combined impact from a number of different factors.
A couple quick thoughts on this.
First, I suspect that similar dynamics apply to many human social issues, maybe especially those that have been called wicked problems. By insistently looking for a single cause or culprit in complex systems at best we fail to address the real problems and at worst exacerbate them.
Second, I think this underscores the importance of robustness or resilience as a factor in the design of all kinds of systems, from datacenters to political and social infrastructure and support systems. A question that we need to ask about all such systems is how stable they are to perturbations or shocks of various kinds. The issue with the bee colonies above is that after the critical threshold has been reached, they are vulnerable to things that would otherwise be sustained without difficulty.