Shrub Thickets of the Future: nature or gardening?

One day Tom Vanderpoel shocked me with a slur. Or so it seemed. 

We'd had a friendly disagreement for years, about the restoration of shrub thickets. I thought it was worth trying. He though it a waste of time – not because it wasn’t desirable, but because it wasn’t possible.

Wild plum - handsome, delicious, and a staple of rich thicket habitat. 

On the day in question, I had expressed some affection, even love perhaps, for certain thickets emerging from our work at Somme. Popular with rare birds, butterflies, and me, these thickets dazzled with plum blossoms in early spring, later pink with Iowa crab, still later delicious with grapes, hazelnuts, nannyberries, raspberries and especially those plums. (For bird details, see Endnote 1.)

Tom said, “Oh, you could maintain thickets. But it would be gardening.” 

What? An insult? Over the years, a variety of people for a variety of reasons have criticized restoration as “just gardening.” They’ve argued that our work wasn’t resulting in real nature, but instead a degraded imitation. Not heroic saviors of the rare, we were mere meddling horticulturists, fiddling around with something too great for us to understand. 

But I knew Tom Vanderpoel. Deeply perceptive, he was not the kind of guy who insulted. He was a smart, plainspoken, generous person. Then he continued, “You’d be cutting out buckthorn forever.”

Ahh, of course. It’s about the difference between restoration and gardening? A successful restoration in time mostly takes care of itself. When you slash, burn, and herbicide the woody invasives out of a prairie, you don’t have to keep doing it forever. The frequent fires of prairie management will kill new seedling buckthorns. Conservative biodiversity is powerful competition; it defeats most invaders in the presence of fire. (But see Endnote 2.)

A burned thicket, a few months after the burn. Shrubs grow back quickly, but they're still smaller than the herbs. In a few years without fire, the diverse shrubs would replace the flowers and grasses, but would something else soon replace them? Thicket species visible above include nannyberry, grape, bur oak, and Hill's oak. 

Thus, as you succeed at restoration, there’s less and less need for gardening-like manipulations. Nature in that rich ecosystem is now on its own. Over the years it will change itself, likely with gradually increasing diversity and conservativism (for a definition of ecological conservatism, see Endnote 3). But those changes will be largely outside the control of us “ecosystem managers,” and we’re glad about that.  

This "nature-takeover magic" can be true for an oak woodland too. When we cut woody invasives from among the oaks, regular burning keeps invaders out. In early years, grass is not the fuel of the oak woods. Oak leaves are. They generate enough heat to kill young buckthorns. (For better or worse, frequent oak woods fires also seem to diminish or eliminate the natural woodland shrubs.) (See Endnote 4.) 

Thus, Tom’s point: Most thickets don't burn. Thus, when you’ve cut the buckthorn out of the thicket, it comes right back from new seeds dropped (pooped actually) by birds. Forever, apparently. In a few years after buckthorn control, new tiny buckthorns start from seed. Year by year, they’ll get bigger, gradually suppressing everything else. A powerful thicket ecosystem does not improve over the years, given the current state of our expertise. We may have to manicure it forever. It would be horticulture. Perhaps for thickets, self-maintaining nature doesn’t get restored. 

Prairies and woodlands differ from thickets in one key way. The former two depend on frequent fire. The thicket is not a high-frequency-and- low-intensity fire-adapted ecosystem. It may be an occasional, catastrophic fire ecosystem. 

Only a few times have I seen controlled burns hot enough to blast a substantial thicket. When they do, the fire does not progress foot by foot or meter by meter, as in grass or leaves. The whole thicket just explodes. In dogwood thickets, hardly any charred wood is left behind. A critical mass is reached that detonates all the wood; an after-image of white powder on the ground is all that’s left amid the blackened grassland where the thicket once stood.

I remember, being taught ecosystem dynamics during a Nature Conservancy retreat at the Archbold Station in central Florida. We stood in an open long-leaf pine savanna and heard how it differed from an adjacent dense tangle. The animals and plants of the long-leaf pine ecosystem are adapted to frequent, low-intensity fire. The big old trees don’t burn. The grass and wildflower plants burn above ground, but roots re-sprout and maintain a relatively stable, conservative turf. 

Those fires consume the grassy fuel up to the edges of the thickets, which, with less fine fuel, haven’t the critical mass for low-intensity fires. Thus, at the thicket edge, the fire stops. But once in a while, every few decades, when by chance a much hotter fire would rage up to the thicket on a drier and windier day, the wood explodes and burns catastrophically. Unlike in the savanna, here the trees die. Their salvation is in their seeds, which sprout only after fire and restore that rare thicket vegetation. Afterwards I wondered, is this Florida thicket in its relationship to long-leaf pines like the relationship back in Illinois between shrub copses and oak savannas?

Illinois rose can make a thicket impenetrable, beautiful, popular with bumblebees, nesting birds, and photographers. The stout thorns make it challenging to control the buckthorn that might ultimately shade out the rose and the whole thicket.

In any case, the problem here in the tallgrass Midwest is that the buckthorn does not respect that long recovery process. It invades, overcomes, kills. It can’t stand frequent fire when young, but with twenty or one hundred years between catastrophes, it recovers fast enough to shade out most of the competition, and the diverse thicket is gone. 

Thus, we don’t know how to restore self-sustainability in thickets. For now, we keep trying. Some animals and plants seem to depend on these thickets (See endnotes 5 and 6), and we want to conserve them. We hope we’re on the road to discovering restoration. Perhaps a diverse shrub thicket will get tight enough to exclude the buckthorn? Perhaps there’s some other route to success. (For recovery experiments, see Endnotes 7 and 8.

Then we’ll have restored thicket nature – as we increasingly do in the prairies and woodlands. For now, in the thickets, we’re partly gardening.

In comments on a draft of this post, Eriko Kojima wrote:

“Even if we think we are gardening, we can’t stop cutting while we try to figure out what to do toward having sustainably restored thickets. While buckthorn, Asian honeysuckle, Malus seiboldii, and multiflora rose are around, we don’t have much of a choice but to cut them out. We do a variety of restoration tasks throughout the year. Excising invasive non-native shrubs to enhance biodiversity of thickets feels like a meaningful, worthwhile, ongoing stewardship task. It’s one of my favorite things. I love doing this work.”

The first draft of this post was published in 2018. 

Endnotes

Endnote 1 

There is a long list of animals and plants that thrive in prairie and savanna thickets, but how many species can’t live without them? Birders have found many species that are eliminated from areas where prairie or woodland restoration has cut out the shrubs. These include brown thrasher, yellow-breasted chat, willow flycatcher, field sparrow, indigo bunting, yellow warbler, and a long list of others. At Somme, our shrub thickets are popular foraging grounds for both Baltimore and orchard orioles, although the orioles typically nest nearby in actual trees. Birds nesting in the Somme thickets include brown thrasher, yellow warbler, yellowthroat, indigo bunting, eastern towhee, and field sparrow. But at Nachusa, original shrub thickets seem to have nesting orchard orioles along with Bell’s vireos, thrashers, kingbirds, and more. Many species take refuge in thorny thickets when danger threatens. They’re also popular habitats for more northern species during migration.

 

Yellow-billed cuckoo. Both cuckoo species appreciate the thickets.
Photo by Jerry Goldner. 

Endnote 2 

We once hoped that fire and competition would eliminate all invasives in time. Some “experts” used to argue strenuously that it certainly would. Perhaps that day will come. But at the current state of the art, we don’t see fire or competition or anything else "natural" eliminating some of the worst. These evil characters include teasel, crown vetch, reed canary grass, sweet clover, and purple loosestrife. We try to wipe them out and watch vigilantly for any new infections. They are like cancer cells, multiplying out of control and killing the host. But most species populations regulate themselves without human manipulation. For example, Somme Prairie Grove now has 490 species of native plants, all of which reproduce and balance each other on their own. Once restoration has established new species, we don’t continue to “garden” them – except in the case of a very few species, mostly those that need protection from over-populated deer.

Endnote 3 

Most species of conservation concern are "conservatives." In a sense, they are the opposite of “pests” and “weeds.” For example, in the “Floristic Quality Assessment” system, plants have numbers (“coefficients” or “C values”) from 0 to 10 that represent their tolerances to degradation and their varying degrees of fidelity to habitat integrity - the higher the number the more conservative:

The coefficient for each species represents the degree to which that species is characteristic of high-quality natural ecosystems. Species given a C value of 0-1 are adapted to severe disturbances, particularly anthropogenic disturbances. Species ranked with a C value of 2-3 are associated with somewhat more stable, though degraded, environments. Those species with coefficients 4-6 occur regularly in good-quality natural communities. Species with C values 7-8 are found mostly in high-quality areas, but can be found persisting where the habitat has been degraded somewhat. Those species with coefficients 9-10 are considered to be restricted to high-quality natural areas.

Adapted from Taft et al. See http://www.conservationresearchinstitute.org/assets/illinoisfqa.pdf

Endnote 4  An Admission of Shrub Failure

It does not work to try to have "a peaceable kingdom" of all species growing on top of each other. As oak woodlands have been restored by thinning, to allow enough light for oak reproduction (and red-headed woodpeckers), they may lose their habitat for wood thrushes as shrubby understories vanish. Thus, efforts to restore or protect shrubs are underway at sites like LaBagh and Somme Woods. One challenge is the fact that the frequent fires required by some woods species often “top kill” the shrubs. Although most shrubs re-sprout readily, they often grow only a few inches a year, keeping them too small to serve as habitat for most birds when managed by frequent fire.

Trying to restore self-sustaining shrub diversity to an oak woods seems not to work, at least in the short run. Under annual assault by deer and fire, the shrubs die out or remain just inches tall. We cage and rake around some small areas. Whether such efforts are heading toward sustainability, perhaps only time will tell, perhaps too much time? 

Endnote 5  

Species of prairie thickets. In addition to the shrubs, some herbaceous plant species seem especially adapted to thickets. These include mullein foxglove, purple vetch, small sundrops, and eared false foxglove. In the case of the last two (both rare), they seem to thrive on the edges of thickets where competition is temporarily reduced when a fire burns into the edge of the thicket. For thicket birds, see Endnote 1. Some rare snakes are said to benefit from thickets, the patchy shade of which helps them with thermo-regulation. Many butterflies appreciate thickets - for example, the great swallowtail, the caterpillars of which eat prickly ash. Very likely, great numbers of beetles, fungi, and all manner of other specialized biota depend on thickets. Tallgrass thickets have been little studied, so far as I can find. Any good references would be appreciated. But the bottom line is that this is a community that we know was once sufficiently common to have many bird species adapted to depend on it. Many other less-studied groups of species certainly lived in that habitat too. Degraded remnants of shrub thickets are not uncommon, as they many of their species have survived in the short run relatively better than prairies in the absence of fire. Much biota may living for now in temporary disturbance situations, waiting for us to reestablish sustainability.

Endnote 6 

Species of woodland thickets. Many woodland birds seem dependent on shrubby understories. These have been said to include wood thrush, veery, Kentucky warbler, and ovenbird. Perhaps some of the shrub-dependent birds were not typical of oak woodlands but more typical of maple, basswood, ash (floodplain), and other types of forest. On the other hand, it seems likely that ravines and wet areas in fire-dependent woodlands would develop shrubby copses. Certainly, that’s the impression left by Baker and Woodruff’s 1910 study. See: The Unexpected Discovery of Somme Woods and Bird Bonanza in 1908. As for other woodland thicket biota, it seems likely diverse, but little studied (much as discussed in Endnote 5).

Endnote 7 

Recovery experiments in prairie and savanna thickets. At Somme Prairie Grove in the early years, we cut most shrubs, following recommendations by Betz and thinking we were restoring prairie. But the “re-discovery of the savanna” required us to go back to the drawing boards. About the same time, birders (especially Judy Pollock and Alan Anderson) were pointing out that prairie restoration was decreasing shrubland bird habitat. We began to search for and restore our more promising shrub areas. Step one was to cut out the major invaders: buckthorn, ash, Tartarian honeysuckle, Japanese crab apple, and multiflora rose. Because we had a lot else keeping us busy, often the buckthorn grew back aggressively.

At the same time, another problem was exacerbated by our efforts. We initially concentrated on hazelnut, because that is the shrub most often mentioned in early accounts of the tallgrass landscape. Because we cut away the buckthorn, our few hazel patches were now more accessible, and the deer ate them mercilessly. They also perversely wracked down larger trunks with their antlers. The hazels soon were smaller than ever. We tried caging a few, which successfully protected them from the deer and gave us a chance to see how they’d respond to just fire. Unfortunately, hazel leaves have a lot in common with oak leaves, in that they are highly flammable. But unlike the oaks, the hazels have thin bark, and even the smallest fire top-kills all. With deer browsing, they’d only grow a foot or so a year, so after forty years, our uncaged hazels are still minor shoots, much smaller than the grasses. Patches we caged and protected from fire for a few years became robust. Now, a healthy patch of hazel may grow five feet tall the first year after a burn. (However, I don't see a lot of birds in them, for what that's worth.)

We’ve also planted many hazels, as nuts and as year-old seedlings grown from the nuts. With caging, many of them survive and grow. But after ten years, for various reasons, none of them add up to much. The best are perhaps two feet wide and four feet tall. A start. But after decades? Perhaps we’re too impatient?

We have begun trying to protect a dozen or so original hazel clumps from fire and deer (a lot of work). These are much more robust than the planted ones. One patch is perhaps eight feet tall, twenty feet long and ten feet wide. That clump is fortunate, for now, in being protected from fire by the fact that no grass or oaks grow close to it. The whole area around it rarely burns much. Very little seems to grow in this big hazel’s shade. No birds have nested in it. (We check for nests every year.)

Wild plum was also commonly mentioned by early tallgrass observers. One plum thicket at Somme has developed well, once we’d stopped cutting enough to notice it. (Plum in winter looks very much like buckthorn, until you’ve learned to recognize it.) This patch is in a relatively wet area that is missed by about half the burns. Unlike with the hazels, this thicket includes many other shrub species including (in rough order of abundance) nannyberry, gray dogwood, silky dogwood, hawthorn, Illinois rose, grape, raspberry, smooth rose, scarlet oak, and bur oak. This thicket is about one hundred feet long and forty feet wide. It’s highly popular with birds for nesting, foraging, singing, and loafing. Nearby is a larger thicket (perhaps 200 feet in diameter) that is also full of birds. It’s still mostly buckthorn, but we’re slowly pushing that devil back to favor the species listed above plus prickly ash, black haw, and hazel. This thicket includes many mid-size oaks, basswood, hop hornbeam, hawthorn, and elm. It’s here that both oriole species nest along with warbling vireos, kingbirds, gnatcatchers, grosbeaks, and many more.   

Decades ago we planted plum pits in one area. They seemed to grow well for a while, but then deer started in on them, and voles and rabbits ate their lower bark in winter. They barely hung on. Two years ago, we finally put hardware cloth around their trunks, and they give the impression that they may finally form a convincing thicket.

More recently, we have also planted many other shrub species at Somme Prairie Grove. They’re mostly still small. But one experiment has been a great success. We threw around a lot of elderberries, and we now have many ten-foot tall elderberry bushes looking hearty – with no other work than throwing those berries.

 

Elderberry bushes six to ten feet tall are emerging in some areas.
All we did was thrown their seeds around. Perhaps the birds brought some too. 

We also protect pussy willow and peach-leaved willow, to see what they can make of themselves. In contrast, we’ve long controlled sandbar willow, which we and many have found to act as an invasive.  

In one case, a pussy willow patch stayed about one foot tall for decades. We surrounded it with a deer-exclusion fence, and it became a fine little thicket. In 2023 we considered it robust enough to make it on its own. We took off the fence, to protect something else, and the result is shown below: 

Buck deer shredded the bark off its stems as they practiced for those antler-pushing contests that so impress the does. How well will this patch do in the 2024 growing season? Time will tell. 

We have planted herbaceous species that seemed right on the edges of some shrub thickets, and some of them end up thriving under and among the shrubs. These include eared false foxglove, mullein foxglove, Culver's root, prairie sundrops, small sundrops, wild coffee, and many, many more. We have not looked for species of darker habitats to see how they'd do under the older, darker thickets, which often have few or no grasses or wildflowers down there in the gloom. Perhaps that would be an interesting experiment. But you'd think that such species would have a way from moving from clump to clump on their own, and we haven't so far seen much indication of them among the nearly 500 native species we've recorded at Somme Prairie Grove so far. One interesting plant that we've found only under dogwood thickets is the fall coralroot orchid. 

Endnote 8 

Recovery experiments in woodland thickets. In Somme Woods the main shrub that had survived in the pre-restoration darkness was black haw. We found perhaps a dozen patches of it, mostly no more than six inches or a foot tall, but covering areas ten or twenty feet across, as if spreading out to search for more light. The largest patch was about one hundred feet across. After thinning the invasives, it seemed to perk up, but after each burn, the new stems would grow only six inches or so, with the result that, given remedial burns every couple of years, they never got more than about a foot tall. Not much of a thicket. In recent years, steward Stephanie Place has raked around that patch prior to burning, and it builds. But this is not quick. It still seems to grow only six inches or so higher every year. Two patches about six feet across are enclosed by fencing, and without deer browsing, they seem gradually on their way toward actual thicket status, now eight feet tall or so.

In Somme Woods, many shrubs including wahoo, gooseberry, raspberry, and Canada plum are found sparingly here and there. Perhaps they’ll thrive with the additional light. Stewards have planted many shrub species mentioned previously along with maple-leaved viburnum and indigo bush. Time will tell how well they’ll do. It’s a lot of work to protect them from deer and fire, but that seems worth it to some Zone Stewards in the short run. One good strategy may be to plant shrubs in wetter areas or areas protected from fire by streams and wetlands. In areas that are likely to burn, it may make sense to plant shrubs more densely in smaller areas, so protecting from fire is easier. Or, are there better experiments, that we haven't thought of yet?

Endnote 9

Note from December 2021. Many people have reported failure in their shrub experiments. I worry about a lot of wasted work. Not every site needs shrubs. Many natural prairies were without shrubs over large areas. Many small prairies today will contribute the most to biodiversity conservation by restoration without shrubs. Large areas of oak savanna and woodland also seem to have historically had few shrubs; it's fine to have few shrubs (or shrubs burned small every couple of years) in some such areas. Some rare birds, invertebrates, plants, reptiles and others thrive in such areas. They're good. Shrub restoration is not something every site needs. Time spent on it may mean less time devoted to some other important need. Diversity in goals and approaches is desirable. 

Comment from Judy Pollock

Tom’s assessment of why our native shrubs are disappearing (birds spreading buckthorn berries) is only part of the explanation. They are also destroyed by deer, drowned by unnatural flooding due to disturbed hydrology, and languishing in the dark of fire-starved woods. It seems to me after lots of experimentation at LaBagh that different shrubs have different stories. For example, red osier dogwood seems too thick and robust to be undone by a buckthorn berry but is a deer favorite and needs more sun than a closing canopy may allow. 

Another “gardening” aspect of restoring the shrub layer is that if you want to have any sort of reasonable density for cover for birds (to replace buckthorn thickets) you have to plant shrubs – which means deciding where to put them, not letting the seeds decide as you would with herbaceous seeds. This requires many hours of exploring shrubs in their natural setting. A very pleasant occupation! 

For all the reasons you describe, we need a thriving native shrub layer, a victim of the lack of fire management and invasive shrubs. I’m glad the restoration community is turning its attention to these important and disappearing plants, even though the customary trick of relying on nature’s magic doesn’t really work to bring them back!

(Judy's valuable points are repeated, along with a response, under "Comments" - below.)