Comment on FR Doc # E8-29671

With respect to: FWS-R8-ES-2008-0006-0028 I here support two aspects of the Critical Habitat designation that have been criticized by other reviewers 1) Incorporation of metapopulation ecology. A ‘metapopulation’ is a set of populations connected by movement of individuals. Within such a metapopulation, populations may become extinct and thereby create ‘empty’ habitat patches that are available for colonization by immigrants. In theory, a set of populations can remain stable over time if the rate of patch colonization balances the rate of extinction, even if each component population is unstable and has a high extinction risk. How is this relevant to Conservation? If a threatened species had completely stable populations, then its conservation could be achieved by protecting a few populations large enough to be immune to inbreeding depression. However, natural populations of most species are not completely stable. Long-term observations and measurements show that, even in systems relatively undisturbed by humans, natural extinctions occur. In this situation, protecting a small number of populations won’t suffice. Instead, it is crucial to conserve networks of habitat patches among which individuals can flow so that natural colonizations have the opportunity to balance the natural extinctions that are bound to occur. Including metapopulation structure into a conservation plan is necessary because it would be from that structure rather than from the persistence of individual populations that long-term stability could be generated. One of the principal messages from a metapopulation approach is that, when metapopulation dynamics apply, “empty” patches of suitable habitat are important if they can form part of a patch network because they are available for the colonization events that are needed to balance out the natural extinctions. Such patches can be prioritized by the likelihood that they would form part of a viable network, e.g.., by their connectivity to occupied patches or to planned restoration sites. Patches that are more isolated and do not form part of a network have a much lower priority for conservation because natural extinctions there are not likely to be 'rescued' by natural colonizations. Does the metapopulation concept apply to Melitaeine butterflies, the subfamily to which Quino checkerspot belongs? Indeed it does! Melitaea cinxia has been a poster-child for the concept and for empirical documentation of the process for more than ten years and figures more largely than any other species in llkka Hanski's (1995) textbook of metapopulation biology. This was the first species for which researchers (again, the Finns) have claimed “metapopulation persistence,” that is, the persistence of metapopulations despite the instability of ALL component populations. Hanski and others have argued that there is no such thing as a stable population of M. cinxia in Finland, yet the species persists due to its metapopulation dynamics. The Finns have censused the butterfly each year for over 15 years across 1600-4000 habitat patches, incorporating the entire range of the butterfly in Finland. There is a wealth of evidence that frequent extinctions and colonizations are characteristic of this species' dynamics in Finland. I'll give two examples of many. First example: The Finns have a paper showing that populations with high heterozgosity are less likely to go extinct. This would not be expected if the extinctions were not real, that is, if 'extinctions' really comprised populations deciding to exist only as diapausing larvae for a few years (like almost any Melitaeine, these larvae are capable of multiple diapause) . Well, actually this couldn't explain the results in the Finnish system anyway because the observations of extinctions and colonizations have been done entirely by censusing diapausing larvae, not at all by censusing adults, so if a population decided not to make adults it would not be recorded as extinct. Second example: Knowing the history of all the populations in the system enables the Finns to sample butterflies from populations known to be young (1-2 years) and populations known to be older (4-10 years). They have measured dispersal ability and its molecular and physiological correlates in insects sampled in this way. Dispersal was measured very directly, by releasing marked butterflies and recording their positions on subsequent days. Among females, the most dispersive insects came from the youngest populations in isolated patches and the least dispersive from the oldest populations in isolated patches. Females in well-connected patches were intermediate. Among males, there was no effect of population age or isolation on dispersal ability. These Finnish results reported in the Hanski et al. in 2004, make excellent sense to me on the hypothesis that isolated patches were colonized by highly dispersive females that mated before they dispersed, so the behavior of the male was irrelevant and there was no selection on male dispersal associated with patch colonization. While a population persisted in an isolated patch, its inhabitants became less and less dispersive because the mobile ones left. On the other hand, if the extinctions and colonizations recorded by the Finns were not real, this result would make no sense at all. What about quino checkerspot? As far as I know no studies address extinction frequencies directly in the manner that the Finns have done, but two sets of work apply a metapopulation approach to other subspecies of Euphydryas editha: Susan Harrison's (1988 American Naturalist) work on the Bay Checkerspot and our own group's work on a quite different set of populations in the Sierra Nevada (Singer & Thomas 1996 American Naturalist, Thomas et al. 1996 American Naturalist, Boughton 1999 Ecology, Boughton 2000 American Naturalist). Our group showed both ecological and evolutionary effects generated by metapopulation dynamics: (1) population density in each habitat patch was partly determined by immigration rates from other patches (2) host preference in a particular patch was also determined by immigration, this time by immigration rates (ie gene flow) from patches where different hosts were used. We found that migration among patches was related to host preference. This has been confirmed in M. cinxia, where the enormous dataset on observed extinctions and colonizations accumulated by the Finns has allowed us (Hanski & Singer 2001 American Naturalist) to show how the colonization of empty patches in the landscape has depended on the match between the host preferences of butterflies that encounter the patch and the host composition of the patch. Since the metapopulation concept DOES apply to M. cinxia and DOES apply to at least some non-quino populations of E. editha, it would be unwise to assume that it does not apply to E. e. quino, in the absence of direct studies of population structure in this subspecies. 2) Predicted range shift in response to climate change E.e. quino shows dramatic changes in abundance from year to year, which are in large part responses to yearly patterns of precipitation and temperature (but mostly precipitation). E. editha as a species is known to respond stongly to climate and so we would expect it to respond also to climate change. Parmesan’s (1996) study showed a significantly higher proportion of E. editha populations persisting at high latitude and high elevation. Because E.e.quino is so dynamic, the high proportion of local population extinctions reported by Parmesan in the early 1990’s may not represete a permanent retrenchment at the southern range limit sof the species. However, there is no reason to expect E. e.quino to respond to ongoing climate change differently from other insects and every reason to expect it to respond similarly to other species that are climate-sensitive. A high proportion of species, not just butterflies, are now showing poleward and upward range shifts. This includes a majority of European butterflies, whose range edges are well-known because of very detailed and organized censuses on a national basis in each country (Parmesan et al. 1999 Nature). Parmesan & Yohe (2003, Nature), found that, overall among plants and animals, about 60 % of species were NOT showing range shifts, but that the other 40% were shifting poleward and/or upward. If indeed climate change is forcing distributional changes this is a problem that conservation biologists must address. E. e. quino may be in particular difficulty because a northward range shift is impeded by the Los Angeles metopolis. There is a current fierce debate within the Conservation community about whether to artificially move species to allow them to cope with climate change. Note that this is NOT a debate about whether climate change is causing distributional changes, it is a debate about what to do about it.