Feb 9th, 2016 by CRF
Nov 18th, 2015 by CRF
CRF Vice President John Laundre wrote the book – Phantoms of the Prairie – on central North American cougar recovery. John predicted cougar recolonization to regions like Nebraska’s Niobrara River Valley. He also suggested that further recolonization to the next good habitat 700 miles east in the Missouri/Arkansas Ozarks will take decades only if state wildlife officials adopt 21st century cougar management protocols. No biologist is more qualified to review LaRue & Nielsen’s Midwest Cougar Viability Study than John.
In the recent study by LaRue and Nielsen, they attempt, through modeling, to predict the population viability of recolonizing cougars in the Midwest. Their model predicts that within 25 years, several potential areas for cougar reestablishment will indeed be inhabited with viable populations. Though this is an interesting modeling exercise and the results appear promising, as the saying goes, what you get out of a model depends on what you put in to it. In this case, it appears that some of the parameters they used leave a lot to question and so too the results.
The first and most important premise they make in the first paragraph, is that recolonization depends on females actually arriving to these areas. Though they state that of the known-sex dispersers, only 76% were subadult males, leaving us to presume that the other 24% were females, they do not clearly state what they consider to be dispersers. Based on the map they provide they show a lot of endpoints around now existing populations in North Dakota and western Nebraska. If we look at the dispersers that have moved beyond these areas, the number of females drops to five the further east/southeast one goes (Bloomington, MN 2002; Grandview, Manitoba, 2004; McCurtain Co, Oklahoma 2006; Tulsa Co. OK 2011, Osage Co. OK 2012). And, although they do mention that their numbers were based on carcass confirmation, they fail to point out that this means most of these points represent almost all DEAD male animals. Instead of a corridor to dispersal, the Midwest is actually a barrier that effectively kills off most who try.
This leads me to the second worrisome point in their analysis, dispersal abilities, especially of females. Their whole model is dependent on data on female dispersal from more western established populations and include an obvious outlier of over 350 km (218 miles). Without this outlier, the average dispersal distance of females is 63 km (39 miles) as opposed to the 112 km (70 miles) they report and use in their model. Two points regarding this: first, the nearest eastern possible population area to an existing cougar population is over 800 km (500 miles) or 13 times further than the average female dispersal distance and second, as mentioned, the further a female cougar moves from an existing population the more likely (>90%) that it will be killed. This intense filter, augmented by hunting (as discussed below), will almost assuredly mean that even if females try to move in a stepping stone fashion, as the authors propose, they will be killed in the extremely hostile stones they try to step on. This accentuates the fact that, the dispersal distances the authors used are from western populations where they are dispersing across often sagebrush dominated rangeland. To equally compare this to the hostile agricultural environment females would not only have to cross but set up a home range and reproduce is tenuous at best, even with the periodic cover of crops such as corn. Two months of crop cover are just not enough for the safety of a female that has to remain in an area for at least 2 years to raise a litter. The bottom line is that the model does not adequately adjust for the harsher dispersal conditions of the Midwest for female dispersal. Again evidenced by the fact that of almost all the known sex dispersers > 50 miles from existing populations, there are just the 5 females noted above, over the 18 years of data the authors present, and they were either killed or captured.
Another problem with this article is that the authors appear to include already colonized areas along the western edge of the Midwest in their analysis of the potential of an area being colonized. I am unclear as to why they would include colonized regions in the Dakotas and western Nebraska in this analysis of potential recolonization. It is somewhat like postdicting that it will rain yesterday after the rainstorm. Of course there is a 100% probably that these areas will be colonized within the next 25 years because they already are! Just like the rain example, to include these areas in their analysis increases the “accuracy” of their model tremendously. It would be interesting to see the same analysis repeated without these pre-existing recolonizations. Also regarding their recolonization predictions, estimated establishment of 1- to 3 females in 4 of the 5 truly current un-recolonized areas within the next quarter century does not provide much assurance that true recolonization of the Midwest will occur in any time soon. And this is assuming the ideal conditions of female dispersal, which I debate above.
The last issue I would like to address is the authors’ conclusion that hunting of cougars will not make a difference regarding the recolonization of the Midwest. This model conclusion seems strange given that the authors state that female dispersal is density dependent. This means that female density has to max out in an area before females disperse. Females are not as territorial as males and so several animals, usually related, will overlap in a given area. It is only when there are enough in the area that a female, probably reluctantly, will move out and then only to the nearest possible place. Given that current hunting of cougars, especially in the Black Hills (the main source population for this eastern recolonization), is aimed at significantly reducing females, I find this incongruent with the model parameter of density dependent female dispersal. Killing of females, especially resident animals, will continually keep the female population below carrying capacity and young females will more likely fill these vacated slots than disperse, especially across the hostile Midwest. On the contrary to the conclusion of the model, only if we DON’T hunt females will an area get saturated enough for female dispersal to occur. What is happening currently in the Midwest supports this contention. In the northern prairie, South Dakota, where they are killing females, little female dispersal beyond the Missouri River has been documented over the last 25 years, so there is no reason for them to leave. In Nebraska, where, but for one year, the cougars are protected, we see more eastern, stepping stone populations showing up, as I predicted in my book and as purported by the model. Contrary to the model output, the real data seem to support the fact that the only way the authors’ model has any degree of accurate predictability is if there is NO hunting of cougars ANYWHERE in the Midwest. Only then can female populations become saturated enough to push individuals to the east.
Overall, given the weakness I have pointed out in this modeling exercise, I, as a cougar biologist, cannot say I have much faith in its optimistic predictions of recolonization. Instead, I see this weakness actually acting as a barrier to their predicted outcome of recolonization. First, it presents too optimistic a picture that recolonizations without reintroductions are sufficient for the return of cougars to the Midwest, and of course, further east. Second, it seems to give the green light to Midwest game agencies for the continued killing of females in existing western prairie populations, and anywhere else across the Midwest. This article will assuredly be used by these agencies to continue the unnecessary killing of cougars and thus further threaten any hopes of recolonization raised by this article.
Dr. John W. Laundré
Sep 27th, 2014 by CRF
Apr 3rd, 2014 by CRF