Rethinking the Ecology of Invasive Species

One Man's Weed is another Man's Wildflower

I just finished reading The New Wild by Journalist Fred Pearce; I found it quite stimulating. As an ecologist, I was thoroughly indoctrinated with the conventional wisdom that exotic species are disruptive, that humans have done terrible things by increasing their global spread, and have thereby accelerated the destruction of Nature. To be sure, I devoutly believed that at the time; after all, I became an ecologist principally because of my desire to save Nature and we all knew there was a an inherent correct balance that humans were throwing out of whack. I was taught, for example, that what we saw as wildlands in my native California were in fact dramatically modified by naturalized species brought in by the early Spanish settlers; saving and protecting what remained and eliminating the invaders we could was seen as paramount. The trouble is, after years of professional study and land management work, I found that I didn't really believe the doomsday narrative surrounding alien species anymore; at least, not in the conventional sense.

In this neck of the woods, the most exemplary bad actor is probably spotted knapweed (Centaurea stoebe), a very aggressive thistle-like invader from Europe. It's actually a rather attractive flower (below) but it has a few ecological tricks that enable it to thrive here. Among these are allelopathy (it produces toxins that deter other species from growing in close proximity), it's apparently unpalatable to wildlife and livestock (though wildlife seems to be adapting), and it tolerates dry conditions (this is a snow-melt area with warm dry conditions common during summers). It is very common here and considerable effort is directed by government agencies at all levels to control it; in my opinion, these efforts are demonstrably failing.

I've always been a little bit suspicious of public noxious weed control programs on public lands because they tend to be reflexive and heavily associated with maintaining forage for livestock grazing, maintenance of tidy roadsides and parklands, etc., as opposed to any clear ecological benefit. But it was when I was called upon to consult with federal regulatory agencies on potential effects of herbicides on native fish that I began to really wonder whether control efforts were worthwhile. Personally, while I don't have much issue with spot treatments of weeds in specific locations (e.g., yards, parks, etc.) widespread application of herbicides to wildland areas simply seems to add the toxins without eliminating the weed.

We have many other weeds besides knapweed, of course. While most people think of plants in particular when they envision weeds, as an ecologist, I was taught to think of any invasive or exotic species as weeds or weedy irrespective of taxonomy. An important invasive (and generally undesirable) vertebrate out here in our streams is the Eastern brook trout (Salvelinus fontinalis). Typically, anglers appreciate them, but they are a true problem for some native fish species, especially bull trout (S. confluentus). These species can be separated within a watershed by somewhat different habitat requirements, but where the co-occur, brook trout have the advantage and tend to replace bull trout. In an ironic twist of nature, brook trout are threatened in some of their native range in eastern North America by an iconic western native, the rainbow trout (Oncorhynchus mykiss). Situations like this tend to be uncomfortable for scientists who continue to believe in the "Balance of Nature" with organisms specifically adapted to their native habitats.

Desirable Non-Native Species

In 2003, the three National Forests in south-central Idaho were among the first Forests nationally to revise their Land and Resource Management Plans (Forest Plans) pursuant to the National Forest Management Act (NFMA). These Forest Plans recognized that some exotic species were desirable components of Forest ecosystems, though they fell short of really identifying which species they were and how they should be managed or considered in proposed management actions; in no case, however, were they considered the ecological equals of native species, and the category seemed to be a management outlier with management emphasis directed at "restoring" natural conditions where possible.

There are a few obvious examples of exotic species that may be considered "desirable." In addition to brook trout, mentioned above, which are widely distributed, popular with anglers, and perfectly innocuous in many areas, we might include: Wild turkeys (Meleagris gallopavo), a sport-hunting favorite introduced for that purpose; California quail (Callipepla californica; pictured), which may somewhat offset the loss (if not helping cause the decline) of native mountain quail (Oreoptyx pictus); and whitetail deer (Odocoileus virginianus), a natural invader from eastern North America that may be adversely affecting native mule deer (O. hemionus) with which they compete and occasionally interbreed.

Then there are Rocky Mountain Elk or Wapiti (Cervus elaphus) and gray wolves (Canis lupus). Both are probably native, but both may have also been reintroduced. The former were undoubtedly native to the Rocky Mountains, but there is suspicion (poorly documented) that their populations may have been very low or even nonexistent in some of this area and some were brought here by train in the 1930s. Wolves, of course, were officially reintroduced in the 1990s, even though some probably still occurred naturally, with wolves brought in from Canada. Elk, of course, are a desirable species very popular with hunters; wolves are much more controversial and can now be hunted legally in Idaho because their reintroduction was, to put it mildly, a resounding success.

Fish are not Plant Taxonomists

 

Fish habitat and population conditions are frequently measured using what amount to one- or two-dimensional indicators, like stream width, average depth, population density (number per unit area) etc. This is sort of inevitable because a reductionist approach is needed to reduce environmental complexity and characterize habitat using readily measured and monitored attributes, but I always thought this sometimes misrepresented reality a little too much. We tried adding some three-dimensional metrics for population status, like biomass per unit volume*, but these didn't help much never really caught outside on our group.

One interesting invasive species is Kentucky bluegrass (Poa pratensis), a common grass native to Europe. We used to encounter it regularly on disturbed streambanks where it seemed to be useful for stabilizing the banks, though there was some concern that it was less effective than native grasses or sedges. In our riparian habitat classification work, we typically recognized bluegrass banks as degraded but somewhat stabilized. I thought that the bluegreass represented an "unnatural" but partially rehabilitated condition and that the bluegrass component might be the "new normal" for such habitats; that is, that the bluegrass was more useful than not because I always believed the three-dimensional structure of the habitat was more important than the species composition.

We explored some of these ideas at a riparian ecosystem workshop in the 1980s with a tentative look at riparian community types on Big Creek, Utah. Classification (habitat typing) of riparian ecosystems was an active area of study at the time as an outgrowth of parallel efforts in land systems classification and vegetation ("habitat type") classification for forest management. These typically centered on the concept of "potential natural condition," but that was clearly a conceptual misfit with our riparian systems because they experience frequent disturbance and successional trajectories seemed to me to be variable. It was then that I started envisioning classification schemes that involved alternative recovery communities that managers might direct, but that effort did not proceed very far.

The Restoration Myth

Watershed, ecosystem, landscape, whatever restoration is a worthy goal; however, it is not achievable. In many cases, the pre-disturbance condition is unknown or poorly known and past-restoration recovery processes cannot be accurately predicted. I think this is true at all scales, from road decommissioning and channel restoration to large dam removal. In my opinion, a better term would be "rehabilitation," which can be used to indicate that some "natural" ecosystem processes are being restored but the end result may be quite different from the original.

An example of undirected rehabilitation we noticed out here was on the South Fork Salmon River. Unwise logging and road construction on shallow soils in steep topography made the watershed vulnerable to a series of heavy rain strorms, including rain on a deep snowpack, and in the winter of 1964-65 it happened. Very little logging has occurred since and and watershed restoration actions, including road decommissioning, have been implemented; naturally this required monitoring and photopoints were established to visually monitor recovery*.

During annual updating of our analysis and reporting on the monitoring effort, I noticed that sometime in the 1980s, about 25 years after the flood events rearranged the stream and deposited thick layers of sand, some of the residual sandbars started sporting some grass. You can see the appearance and growth of this grass in this short video that I made:

We looked into this a little and discovered that the grass was reed canarygrass (Phalaris arundinacea), a plant whose origins seem to be somewhat debatable, probably from Eurasia but possibly with some native new world populations, and often considered invasive or undesirable when in an inappropriate environmental context. Discussions with some watershed rehabilitation professionals led us to conclude that, although it was never used in post-flood recovery work in the area shown and we caouldn't determine whether it existed naturally, it probably was used farther upstream to help stabilize streams. It is obviously working to that end here, quite successfully, and although one major criticism of the species is that it forms monocultures that inhibit other species, we have noted willows beginning to form on the pictured sandbar.

This demonstrates the rehabilitation of a heavily damaged riparian system resuming effectively natural function with a suite of species that includes some that were probably not naturally present. The river has regained much of its pre-development character and resilience, but is still different than it was naturally; it has been rehabilitated, not restored.

 

A "Corrupted" Gene Pool is better than no Gene Pool

 

Fisheries scientists seem to have always had a different view of genetics on preserving wild populations than wildlife biologists have. This is just a personal, subjective observation on my part, but I think the evidence supports it. We have had wolves re-introduced out here from populations that were native to Canada. Surely they are the same species, but all sorts of things can vary among populations within species. Wolves were not extinct in Idaho when new ones were introduced, but packs (seen as a fundamental characteristic of healthy wolf populations) were apparently uncommon. So far as I know, there were no genetic or behavioral studies of Idaho and Canadian wolves investigating their similarity or comparability and building wolf packs was seemingly the principle goal; it is an understatement to say that the program was a success in that respect.

I'm not an expert in wolf biology and offer the above synopsis mainly to contrast it with the situation with Pacific salmon. Although I have heard from people who have expressed concern that the introduced wolves were behaviorally different from native Idaho wolves with the former displacing the latter, I know of no rigorous study of the issue. With Pacific salmon, however, there have been concerted efforts to prevent mixing of populations among watersheds, even to the point of turning fish away from spawning areas if they had strayed from a different population***. Because salmon runs in Idaho are much depressed over natural levels, largely due to the massive hydroelectic dams on the Columbia and Snake Rivers, supplementation with hatchery fish is improtant; the uptick in salmon redds on this chart from the South Fork Salmon River may reflect this importance. In this case, salmon collected in

the South Fork Salmon River and reared in the McCall hatchery are planted in the South Fork Salmon River; they are often back-crossed with wild fish to reduce the potential for decreasing fitness from artificial breeding.

Unfortunately for people who wish populations would not intermingle, salmon are incompletely faithful to their natal spawning grounds. Thus the situation mentioned above with salmon going to the "wrong" weir happens with some frequency on spawning grounds. From a biological perspective, this is probably "good" in the sense that it allows fish to disperse into new habitats.  So mixing will occur naturally regardless of whether we wish it. In the South Fork Salmon River, the hatchery fish are fin clipped (their adipose fins are removed) for easy recognition of their non-wild status. The Secesh River, a major tributary of the South Fork Salmon River receives no stocking of hatchery fish and is looked upon as a wild fishery, but hatchery strays are likely inevitable. Here is a photo I took of two Chinook attempting to spawn in a tributary of the Secesh River; it's difficult to interpret, but I believe the upper one is missing its adipose fin.

Parting Thoughts

I think care is needed when introducing exotic species, and I think that targeting specific instances where invaders can be stopped and trying to stop them is reasonable. I also think it is reasonable for rehabilitation actions to construct viable recovered states that are not necessarily intended to duplicate natural systems. In the long run, however, Nature will likely be able to take care of itself.

___________

* This stream canopy paper is an example.
** This was just part of an extensive monitoring program that has led to numerous reviews, reports, and even peer reviewed publications for about 50 years of post-flood monitoring, one of the longest continuous data collection efforts in existence.
*** I cannot personally verify this, but it was told to me by a biologist friend whose job it was to do such things.