Tuesday, April 25, 2023

Good Questions about How to Burn

from Drew Harry of Madison Audubon's Faville Grove Sanctuary in Wisconsin




This caveat may be boring. But it’s important. No one knows the “right” answers to ecosystem “prescription” questions. That’s partly because this “science” or “medicine” or “physical therapy for the ecosystem” is relatively new. It’s partly because every site and natural community is different. It’s partly because the way we learn is for many of us to try differing approaches, keep records, and compare. And yet – we have responsibility for “life and death decisions” about irreplaceable remnant ecosystems. We have to decide and act.

Q and A  

Q: I'm wondering how you all handle raking around trees before woodland or savanna burns.


A: The answers differ from area to area and from tree to tree. But a central principle is that we hope the fire will kill many trees. Too much tree shade is the greatest threat to oak woodland biodiversity. Bur oak and white oak woodlands are two of the mid-continent’s most endangered or threatened ecosystems. The threat to them is not too much fire … but the lack of it. 


Most woodland burns, while "top-killing" buckthorns and some other small trees, do not reach temperatures hot enough to damage most trees, including the ones that merit control. 

In a bur oak woodland, excess shade from invading tree species of many kinds kills off the lower oak limbs (leading to rot) and prevents reproduction (as oaks need more light than the invaders). Many bur oak woodlands have had no bur oak reproduction for over one hundred years. Although it doesn't often happen, we'd hope our burns would be hot enough to kill most other trees. In the past a natural bur oak woodland was often mostly bur oak - with a few other tree species here and there. But in the decades without fire, those other species have grown so numerous and dense as to become a kind of ecosystem pathology. We invest a lot of stewardship energy in sawing them down and burning them in bonfires to let in enough light for bur oak reproduction and for the welfare of all the other plants and animals of the now-rare bur oak woodland ecosystem. To the extent that fire will do some of this work, so much the better. 

In this degraded oak woodland, the biggest, oldest trees are bur and white oaks. Neither have reproduced in recent decades. Aside from a few red oaks, the principal new trees are invading maples. For biodiversity conservation, we would hope for fire intense enough to kill most of those maples and restore enough light for reproduction of the bur and white oaks. 

Only a bit less threatened, white oak woodlands in our experience have much less fuel and much tamer fires. But the principle is similar. Especially in areas of good soil, invasive maples, basswoods, cherries, elms, red oaks, and other trees have already become too dense to allow white oak reproduction. As we cull many of them by saw, girdling, and fire, literally hundreds of species of rare plants and animals are able to return as light levels increase. 


This hickory has been scarred by fire four times. 

On the second such occasion, the heat almost girdled the entire circumference, which would have killed the tree (about 30 years old, at that time). But after each fire, enough phloem and cambium remained that the tree recovered. This year, following 50+ years of survival, it succumbed to our saw and that purple herbicide. Hickories are good. But this woods had far too many of them.  


Leaf litter fires and the resulting scarring of trees are parts of nature, and the process favors bur and white oak trees which are keystone species to two of our most endangered ecosystems.


We also protect many young bur and white oaks and some shrub thickets by raking and backfiring. We only slowly became conscious of the fact that we needed to protect bur oaks at Somme Prairie Grove if we were to see reproduction of actual trees. All the young oaks in open areas, because of burning since 1980, were re-sprout bushes. 


Do you let headfires run through the leaf litter, or patiently wait for the backburn?


We burn with headfires as much as we can. They’re hotter and do more of the work that we want fires to do. Other people give good reasons for preferring backfires. It would be great to have long term studies comparing the two approaches. But as a default position, if we’re emulating nature, most land burned by headfires, as they moved the fastest and covered most ground. 


How do you handle the prairie/savanna margin? Do you let a hot fire run into the savanna?


Yes. In such situations we follow the principal “Let the fire decide.” In our experiences, we haven’t found “too hot” fires to be a problem. On one site where I work, we had an artificial situation where dense prairie grass grew right up to the edge of planted white oaks, then perhaps 20” DBH. The fires over the years killed most of those mature oaks. And yet, the site was former bur oak savanna. We were happy to see the planted white oaks retreat and the burs advance. Possibly, it once was the natural order of things for bur oaks to abut the prairies. White oaks seem to have grown in somewhat more protected areas, indeed protected in part by the reduced fire under the bur oaks between them and the prairie.  


Fire in the open grassland typically has flames many times longer than fire in the woods. 

The tree on the left is a Hill's oak (in the area shown above). Given its thin bark, much of the lower part of this tree has been killed by the previous fire. On the right is a young bur, with much thicker bark, little impacted by the burn. Over years of burning, this bur oak may outcompete the Hill's on this open grassland edge. 


And do you change your actions based on humidity, temperature, wind etc.?


A good answer here might take a whole book. So perhaps, for now: Yes, certainly, in so many ways. 

Do you still burn when spring ephemerals and other spring flowers are blooming? The leaves never seem to dry out by the time hepaticas and bloodroot start blooming, but they never seem to fall and dry out before snow in late fall/early winter either. 


Yes. We sometimes burn plants that are evergreen - or stay green late in fall - or green up early in spring. Last year's evergreen leaves are not the ones that will do the main photosynthesis in the new year. Those species that are part of the oak woodland ecosystem, which are all or most species that we find there, come right back. The spring flora is precious … thought-provoking … refreshing … especially coming, as it does, after the barrenness of winter. It’s emotionally hard to hurt that spring flora. 


As you point out, Drew, it’s similar in the fall. Many species are now so rare and precious that it’s emotionally difficult to burn gentians and asters while they're in flower. But fire-adapted plants roll with the punches. It’s good for the oak woods ecosystem to get a good burn and ultimately good for spring and fall species too.  If they must skip a year of seed production, that’s a small price to pay for overall health of the ecosystem they depend on. 

That being said, it's possible repeated late spring burns when wildflowers are up may lead to declines in some of those species – though I’m not aware if anyone has truly studied that. Increased competition from recovering summer and fall species may also conceivably decrease some spring species that have profited from lack of such competition. There’s much to learn.

How do you handle downed wood?


With such an excess of dying trees (from ash and elm disease as well as from fire), there can be unnatural amounts of fallen trunks and large branches. We mostly hope it catches fire and burns up. Some managers carefully rake around dead wood … and then extinguish the logs that catch fire anyway. But that means they must take the time to do that same work year after year – meaning less time for work that’s much more important to both safety and the ecosystem. 

In some areas we see thick accumulations of unburned decaying smaller sticks, bark, twigs, nut shells, and other litter. Ray Schulenberg from the Morton Arboretum observed that such rotting woody material acidified the soil or otherwise negatively impacted soil chemistry. We continue to hope that fires will be hot enough to burn that up.

We pull logs and dense branches away from oak trunks (both old trees and reproduction, if any) to the extent that we find time for it.  

A low-intensity leaf fire can kindle a much hotter wood fire where a downed log lies close to a tree trunk.  This dead log against a red oak trunk led to a fire that damaged the red oak (and continues to consume the downed log days later). Protecting the young white oaks (pale barked trees in the middle distance) was a priority here. It we'd had time, we would have moved a section of such a trunk that was up against a white oak. It did not concern us that this red oak trunk was damaged. Indeed, for biodiversity conservation purposes, we are culling red oaks in this area.

Fire between this trunk and log badly burned this bur oak trunk years ago and re-burned a part of it recently. In a woods or savanna with little bur oak reproduction, if we saw it ahead of time, we would have cut and moved the part of the log away from the tree. 

How does the size of a unit affect the care you're giving to individual trees?


Here I can’t resist invoking one of my favorite burns. At Nachusa Grassland we burned a circle around about 400 acres of prairie, savanna, and woodland on a spring day. We had no prepared firebreaks, as we knew where natural firebreaks occurred and where elsewhere we needed to go slow with special care. During the burn, Wilson’s snipe winnowed over the wetland and wild turkeys gobbled in the woods, while surrounded by flames. Ducks landed in the ponds during the burn, perhaps expecting the fires to drive tasty insects to them. In the evening, now with 360 degrees of burned firebreaks, woodcock performed their mating rituals as the fire burned up all it wanted within that circle, wandering idly in some places, then flaming dramatically in others, until well after dark. That was decades ago, before any of us had heard of Nomex flame-retardant clothing. We used no vehicles – just legs, arms, flappers, backpack sprayers, and rakes instead of drip-torches. 


Yes, large burns are different. We typically don’t have time to worry about details within the burn. The fire decides. 


Because they nest on the ground in early in spring, we often burn up woodhen’s nests. We are sorry. If we run across a nest by chance, we rake around it and spray water over the area, and we have seen such nests survive. But some are burned. The woodcocks will nest again.  

Is this bad for woodcocks? Our fires restore such good woodcock habitat that our bird reports regularly get skeptical questioning from Ebird. Woodcocks have in recent years spread out of the savanna and into the now-more-open and thriving oak woodlands. Belatedly we now understand why they’re called “woodcocks.” 

Taking the long view, oak woodland restoration has been very good to birds. But you have to break some eggs to restore good habitat for all. 


Drew's questions and this post refer back to this burn report.

Additional References:








Thanks to Drew Harry of Madison Audubon Society for good questions and discussion.

Thanks for proofing and edits to Christos Economou and Eriko Kojima. 


  1. With respect to headfiring and backfiring, and with particular respect to white oak-dominated sites and woodlands that aren't so purely oak. There are certain places on the landscape where fire would almost certainly have been mostly backing (in S WI these are often white oak sites), often topographic breaks (often N and E aspects, but not always depending on where water is) where fire most often would have been going downhill. But that's very site specific. A caution I make urgently where topography is steep and oaks are mostly not mature bur oaks is to be very careful with uphill headfires especially where leaf litter and debris have accumulated excessively with fire exclusion, because I've seen a few instances where uphill headfiring has caused mortality of most mature non-bur oaks in a stand. Again, this is all very site specific. Sugar River savanna near Madison has been burned all years but three over the past 45+ years, and there was a 2 year hiatus (year before covid and then burn permits shut down for covid) that allowed old white grubs to make huge gains, and upon resumption of fire, enough oaks after only two years were beyond the grub state that additional action was needed to prevent excessive oak regruitment--so the more intense the fire there, perhaps the better.

    1. I often say, “fire is not a panacea.” It is a tool. We need to make sure it is the right tool for the job. Once established, many woody species will not be killed by fire. The invasive common buckthorn and Asian bush honeysuckles all sprout after fire, at least this is true if they have reached a certain size. In contrast, fire repeatedly top killing native shrubs can eliminate these shrubs. Often herbicide is the best tool for the job. My observation has been most herbicide applicators don’t know how much needs to be applied to kill the invasive species they are targeting. When herbicide applicators are successful in controlling invasive species, they almost invariably do it in a manner that kills many other things defeating the whole purpose. Like a burn boss, I think we need a position specific to trialing herbicide application techniques, and monitoring herbicide applications for quality to work toward continuous improvement.

      Not everywhere will burn. Inside the curves in stream meanders in woodlands don’t burn. These areas tend to be a haven for invasive moneywort. The Forest Preserves of Cook County burns the Shoe Factory Road Nature Preserve every year. However, despite removing almost every tree other than burr oak, the east side of this hill never burns. I saw one hackberry (thank you butterfly monitors), but otherwise just burr oaks. If my memory from many years past is correct, the thinning has benefited the ground layer. Pennsylvania sedge and woodland wildflowers have spread. I have some pictures on the following blog post.


      This woodland benefited from thinning and careful application of herbicide. Despite fuel and annual burning, the fire never carries through the east side of this kame.

    2. Thanks to Dan Carter for good comments from a wise and careful observer.

    3. Kirk Garanflo sent me an email with a photo showing half of the eastern slope of Shoe Factory Road Nature Preserve having been burned. Mr. Garanflo explained that the other half was not burned because of downed logs. The FPCC did not want to get these logs ignited and have them burn throughout the night near a road.

      I helped burn the Shoe Factory Road Nature Preserve in the past. The fire died as soon as it hit the east facing slope. I have watched this preserve for many years, seeing where the fire had stopped when it got to the borders of the eastern slope. After years of management by the Poplar Creek Prairie Stewards, it must only have been in the last few years that fuel species have recovered to the point fire would carry. I must have missed this more recent turn of events. During my recent visit, I saw plenty of fuel on this slope. I will watch in the future to see if fire is able to carry through the eastern slope of this preserve when prescribed burns are conducted.