Tom's Blog

Wednesday, April 25, 2018

Finally: Good weather for controlled burns

We’ve been burning so much this spring under marginalweather conditions,  that it was a surprise yesterday to have ideal weather for our burns. Not only did we complete our regularly planned burns, but we were able to re-burn many of the units that had not burned well in the February and March burns. As Amanda said: “The day everything burned REALLY well”.  At 10 AM the temperature was over 50 F and the R.H. was 42%. By mid afternoon it was 70 F with RH around 30%.

Of course it was April 24, very late for us to be burning, but you have to take what you are given. Fortunately (?), the very cold April kept almost all of the plants underground, so they were protected from fires. The few species that had already shown a few shoots above ground should have no trouble re-growing from dormant root-stock.

Why do we burn with marginal weather? Most of our burns involve oak savannas or woodlands where burn conditions are often iffy. However, it is so important to get these burns done that we do them when we can. The idea is that we can create reasonable peripheral black lines and get as much of the interior burned as possible. Then, if better conditions arise later, we can go back and re-light areas that did not burn well. And this is what we did yesterday.

Unit 10 savanna burn of Feb 27-2018. Although black, there is a lot of unburned thatch. 

The same savanna after the re-burn on Apr 24, 2018.
Complete combustion, as shown by the white ash

In contrast to savannas, prairie burns can generally be very successful in marginal weather conditions, provided the fuel is fully cured and the dead fuel moisture low enough. The same day that our savanna burn was not very successful, the nearby prairie burned very well.

No thatch left after this burn of Feb 27, 2018

Wednesday, April 18, 2018

Costs for restoration of savanna and prairie remnants

Restoration of oak savannas and prairie remnants is usually a complicated and expensive process, but real data on costs are hard to find. From time to time during 20 years of restoration work at Pleasant Valley Conservancy I was able to record times involved in various tasks. Since most work involved multiple workers, I have converted the time sheets into person/hours. Areas restored during the same activities were measured either by walking the periphery with a Garmin GPS or by GIS using air photos with ArcMap.

The various work activities shown in the summary table need explaining. Most of the work considered in this post involves removal of woody vegetation, both trees and shrubs. Removal of herbaceous plants, weeding, is a different activity and is not presented. Photos are added to present the context of the restoration activities.

The procedures discussed here have been used on oak savanna or prairie remnants that are “restorable”. That means that they once were “good” remnants but had become seriously degraded because of grazing or lack of fire. They still had some native species and retained small remnant areas.

Basal bark herbicide treatment
One of the most efficient ways of eliminating woody vegetation is basal bark treatment with an oil-soluble herbicide that will penetrate the bark and kill the roots. Since basal bark treatment can be done any time of the year, it is often done in the winter when many other restoration tasks cannot be done. At PVC, 100 foot swaths were marked with flagging tape and workers move along these swaths, treating any target species. The treated plants died during the next growing season and the dead plants were left in place. In a year or two they fell over and slowly decomposed. This works best for scattered plants.

Cut and treat
Denser populations of brush such as honeysuckle, buckthorn, or brambles, were cut with motorized brush cutters, followed by cut stem treatment with herbicide. Depending on the species cut, glyphosate or triclopyr was used to treat the cut stems. This procedure involved a team of people, one operating the brush cutter and three or four following with herbicide spray bottles. This procedure was used primarily on infestations of smaller shrub heights (2-4 feet), and the cut brush was just left on the ground to burn up at the next fire.

Tree work
Many restorations involved the removal of invasive trees such as walnut, elm, cherry, and box elder as well as brush. Often these trees were 10-20 inches in diameter and an experienced crew (contractor) was needed. If the trees were smaller, so that removal was less technical, an experienced chain saw operator could work with a group of less experienced workers. Tree work requires much more time because brush piles must be built, and the cut logs must be removed from the site.

Foliar spray (reprouts after a burn)
A very efficient method of removing brush is to burn the unit (either spring or fall). A properly conducted burn will top-kill the shrubs. In late spring or early summer new shoots will appear and can be sprayed with an herbicide such as aqueous triclopyr. This is a very effective procedure and is less time-consuming than others. There is no cut wood to deal with, and the sprayed plant is usually eradicated.

Foliar spray (fall)
In the fall, buckthorn and honeysuckle retain their green leaves long after all the native vegetation has senesced. Although the timing is variable from year to year, these plants can usually be killed by a foliar spray such as glyphosate or triclopyr. Remember: if it is green, it will be killed. Senesced plants (brown) will not take up herbicide. Glyphosate is preferred because there is no soil residue.

Examples of the techniques  used are given at the end of this post.

The Table
The table is arranged in two ways
  •      By task
  •        By hours per acre from low to high

I used hours per acre to express costs because hourly rates vary widely,. However, I have added a column assuming $25 per hour.

Wooded road bank
Basal bark
Mixed brush
Oak woodland (whole North Woods)
Basal bark
Large buckthorn and honeysuckle
White Oak Savanna (Unit 12A)
Cut and treat
Moderately steep; brambles; honeysuckle; no buckthorn
South slope and basin savanna; 2006
Cut and treat
Mixed small brush; steep terrain
South slope and basin savanna; 2007
Cut and treat
Mixed small brush; steep terrain; snow year
Bur oak open savanna (Unit 11A)
Cut and treat
Brambles and mixed brush
White oak savanna on steep slope
Tree work (contractor)
Many large walnuts to cut
South slope below diagonal trail 1998
Tree work (hourly workers)
Moderate-sized trees and big brush; chipping
Ridge top and basin savanna

Tree work (contractor)
Trees and brush; lots of wood to deal with
East Basin
Tree work (contractor)
Fully closed forest converted to prairie
Bur oak savanna (Unit 8;10)
Foliar spray
Brambles; brush resprouts after spring burn
Bur oak savanna (Unit 10) 2007
Foliar spray
Buckthorn; fall foliar spray
Bur oak savanna (Unit 10) 2008
Foliar spray
Buckthorn; fall foliar spray
East Basin
Foliar spray
Spot spray after clearing and burning

$$ per acre based on $25 hourly rate; other rates are possible, or none, if the work is done by volunteers.

Basal bark is one of the most economical methods of brush control. However, the dead brush is left standing, and one must wait a year or two for the dead stems to disappear. However, it could be used as a preamble to a major tree cutting operation. If it was known that a lot of trees were going to be cut in a unit in a year or two, it would be efficient to basal bark all the woody vegetation. The dead wood could then be used as the base of the brush piles that would be made from the smaller branches of cut trees.

Foliar spray of resprouts after burns is another economical method of brush control. By spraying the resprouts before they get big, one is eradicating the plant and preventing it from getting larger. The principal problem is timing. Not all plants resprout simultaneously, so that it is necessary to canvas each area at least twice, and preferably three times. Also, spraying usually must be done in May when herbaceous weed control is also underway.

Cut and treat on a honeysuckle-infested hillside. One brush cutter and three treaters

Cutting and treating in winter. Light snow is helpful in keeping track of where the brush cutter has been.

Cutting and treating near the Town road makes it easier to deal with the wood.
This is the beginning of the major job to clear the whole South Slope.
Hourly workers were used rather than a contractor

This area had many large trees that had to be cut. Contractor

Bramble resprouts after top-killing from a burn.
This is the ideal stage for foliar spraying

Spraying resprouts. It takes a good eye to locate all the resprouts.
All woody resprouts are sprayed.
A 3-4 person crew is ideal.

Buckthorn green. All the native vegetation has already senesced.
An ideal time for foliar spraying

Four-person crew canvassing a savanna for green buckthorn.
A systematic search of the whole unit is necessary, using marked swaths.

Saturday, April 14, 2018

An interesting response of sumac to periodic burns

Sumac can often be a serious invader in prairies and savannas. Although it is regularly top-killed by fire, it reprouts immediately. In addition, sumac root suckers readily, so fire can often appear to stimulate sumac growth.

Recently I learned of a long-term fire study in oak woodlands of the Missouri Ozarks where sumac greatly increased with a four-year burn interval, whereas there was no sumac in unburned or annually burned plots. What is going on here?

The study area is at the University Forest Conservation Area in Butler County in southeastern Missouri. There were three treatments: unburned, burned annually, and burned on a 4-year interval. This study began in the early 1950s and has been continued up to the present (see reference).

Knapp and co-workers made extensive measurements of the understory and overstory in these experimental woodlands. There were a number of interesting observations, but the focus here is on  sumac growth.

Neither the unburned controls or the annually burned plots had any significant sumac.However, sumac was the dominant shrub in the plots burned every four  years (see photo). The species of sumac was Rhus copollina.

Photo courtesy of Benjamin O. Kanpp

How should we explain this?

First, it is not surprising that the unburned plots have no sumac, since sumac does not grow in shade.

Second, it is not surprising that the plots burned annually do not have any sumac, since it would be top-killed by fire. In fact, there is virtually no understory shrub layer in the plots burned annually.

But how do we explain sumac thriving in the plots burned every four  years? Most likely, the fire is unable to carry through the sumac thickets. The stem density of a sumac clone is high, and the leaves will prevent oak leaves, the principal fuel, from reaching the ground.

Stem density of  a sumac clone. The ground is essentially bare. Fire would probably not carry through

It appears that once sumac gets started, there is no stopping it!

This is not the first time I have heard about 4-year burn cycles not being sufficient to control brush. See the paper from Konza Prairie by Briggs, cited below.

Knapp, B.O., Stephan, K. and Hubbart, J.A. 2015. Structure and composition of an oak-hickory forest after over 60 years of repeated prescribed  burning in Missouri, U.S.A.. Forest Ecology and Management 344 (2015) 95-109.

.Briggs, J. M., A. K. Knapp, and B. L. Brock. 2002. Expansion of woody plants in tallgrass prairie: a fifteen-year study of fire and fire-grazing interactions. American Midland Naturalist 147:287–294

Tuesday, March 20, 2018

Prescribed burning in cold weather

This season we have been doing prescribed burns in fairly cold weather. The dew points have been quite low, so that there was no dew on the grass even when the temperature dropped to the mid 20s (F).

Leading up to each burn day we had several days of clear, sunny weather with temperatures reaching the upper 40s. But for the two successful prairie burns we did on Friday March 16, 2018 the temperature never got above 32 F! We burned the East Basin (5 acres, with complicated topography) and the Valley Prairie (4 acres adjacent to the PVC wetland).

I had no success finding any information in the fire literature about burns in cold weather. Perhaps most people think you can’t burn when it is cold.

One of the most important parameters for a prescribed burn is dead fuel moisture content. This time of year prairie grass is fully cured, which means that its moisture content is completely controlled by external conditions, most importantly rainfall (or lack of it) and relative humidity (R.H.). Although temperature does have an effect on the rate at which moisture of the grass is taken up or lost, the most important influence of temperature on dead fuel moisture is the effect on R.H. When we burned Friday, the R.H. was below 50%.

Although the fire literature has tables for estimating dead fuel moisture, these are only approximations and don’t take local conditions into consideration. But there is a simple way of estimating dead fuel moisture of oak leaves, which are a fairly good surrogate for grass.

This is McCarthy’s test for assessing the moisture content of hardwood leaves by bending.

Moisture content
Behavior during bending
Leaves crack if creased but do not break entirely
Leaves crack if folded more than a right angle
Leaves crack if bent at a right angle but do not break freely, especially at the veins
Leaves break entirely apart if bent at right angles
below 10%
Leaf crumbles when you pick it up

Source: McCarthy, E.F. 1927. Weather and forest inflammability in the southern Appalachians.
Monthly Weather Review, March 1927.

 Here is a guide to using the McCarthy test: When using the bend test, take several samples in both sunlight and shade and if on a slope, at the top, middle, and near the bottom of the unit. If the leaves have been wet from precipitation, wait at least one day of full sunshine with medium winds and RH below 50%.

Just before we started the East Basin burn, I used the oak leaf bend test. It gave a reading of 14%, which is a good dead fuel moisture content for a prairie burn

The photos show the fire behavior in the East Basin Prairie, a restored prairie. The fuel here, mainly Indian grass and little bluestem, was in its 8th growing season and obviously well established.

According to NOAA weather radio, the early afternoon temperature was 32 or 33 F over all of southern Wisconsin.

Successful prairie burn under cold-weather conditions

Keeping a straight burn line on  the east side of the unit

Doing some interior lighting on a later stage of the burn
This prairie has a complicated shape, with level areas surrounding
a steep south-facing basin

Two fire lines meeting on the final stage of the burn

Almost complete burn coverage with just a few unburned patches.
The white ash indicates complete combustion.

The south fire line (facing north) still had piles of snow.
Note also the snow on the distant hill.

Final check of the burn.
Most of the small smokers are well inside the unit and will be left to burn up

6146 6157 6163 6165 6168

Tuesday, March 6, 2018

A very successful early-season oak savanna burn

The Weather Bureau made no mention of it, but March 2-4 2018 provided unusually good weather for controlled burns, especially for oak savannas. This is quite a bit earlier than we usually do savanna burns, but one element of burn planning is to burn whenever the weather is right.

Carrying out controlled burns of natural areas require lots of attention to detail.

We had already done a major burn of our South Slope prairie and bur oak savanna remnants on February 27, with the same kind of weather conditions. After a few days of uncertain weather, the good weather returned. A cold air mass moved in with low dew point (21 F) and moderately warm temperatures (45-50 F). Especially important was that the sky was clear and the sun was strong. (In early March the angle of the sun is already high enough to bring on significant warming of the fuel;  grass and oak leaves.)

As long as you have low R.H. and sun, you don’t really need warm weather, provided it is above freezing. An additional important weather element was a good wind (8-10 mph) from the south. (Because oak savannas often have downed woody debris, a good wind is an important element for a savanna burn, since it helps to keep the fire moving along.)

It took a day to get everything ready, so the burn was scheduled for March 4, a Sunday, giving us the opportunity of getting more than usual volunteers. These together with the Integrated Restorations crew gave us 16 people, all experienced burners. (The IR folks understand about burn scheduling and are willing to work on Sunday!)
The crew assembling at the anchor point. The vehicles were parked on a black area from a previous burn.
Each person had a two-way radio, essential on such a burn.
We burned all the area shown in light tan on the map, 36 acres. Because of logistics, those units north of the motor vehicle road (Unit 19B,C,D, and E) were burned separately after lunch along with several units from a future planned burn (Toby’s Prairie, Annex, Triangle; see map); 6 acres.

We call the units we burned the Basin Savanna, since they essentially wrap around the Pocket Prairie, which acts like a “basin”. The oaks in the upper part of the basin are burs and those lower down predominantly white, with some Hill’s oak and black oak intermixed.

The location of the start of the burn, which was also our anchor point, was the junction on the map where 19E, 12B, and Toby’s Prairie come together. One crew moved west from there along the gravel road, backburning down through 11B and 11A until they reached Unit 18. The other crew moved east, backburning through 12A and  20 until they reached the Pocket Prairie.

Timing was critical and it was essential that the two crews coordinate. We did not want the 2nd crew to get too far ahead because once the 12A fire line reached the Pocket Prairie it would touch off a head fire in the prairie that could sweep up into 18 and 11D. Thus, the 1st crew had to get the 11A backburn finished first. Fortunately, every crew member had a two-way radio, so that coordination was good. Amanda, the burn boss, monitored the two crews carefully.

The Basin Savanna burned very well, benefiting from the RH of 35-40% and a strong drying wind gusting from the south or east.

We had twelve drip torches, and used every one!

One of the important requirements of the savanna burn was to protect the lower branches on the large trees, especially those attractive “savanna oaks”. The way to do this was to avoid head fires in the savanna.

Another goal was to protect as much as possible the living birches, since they are greatly used as wildlife habitat (woodpeckers, bluebirds). The litter at the bases of the birches had been removed with a leaf blower.

Another goal was to protect snags (great wildlife habitat) by clearing the litter around them.

Another reason for backburning is that it keeps the flame height low.

However, because of the gusty wind, random change from backfire to headfire sometimes occurred. Although the wind helped to carry the fire in the savanna, it had potential danger for special trees, birches, snags.

Tree-centered spot firing One procedure for protecting special trees, birches, and snags, was tree-centered spot firing. The drip torch operators had been instructed in how to use this technique. See my blog post for details.

The savanna units back-burned well but gusts occasionally turned the fire into flank or head fire in some areas. Fortunately, these were relatively infrequent.
Sometimes all one had to do was watch the savanna burn. This is a late stage of the burn. Ron Endres photo
A view across the Pocket Prairie to Unit 11D of the Basin Savanna. The oaks at the top are burs, those below are whites.
The fire line moving down through the Basin Savanna eventually reached the Pocket Prairie, jumped the firebreak, and started the Pocket Prairie on fire. The Pocket Prairie mostly burned that way but was also lit from Pleasant Valley Road. Fire also started in the SW corner of the Pocket Prairie area by a spot fire from Unit 18. 

One of the goals was to get the fire into  Unit 11A and Unit 18 before the Pocket Prairie caught on fire, because when the Pocket Prairie burned there was a possibility of it igniting Unit 18 from the bottom and hence turning Unit 18 burn into a head fire.

Note that with the lush fuel of the Pocket Prairie and the low RH, the Pocket Prairie burns very rapidly. It takes no more than 10 minutes for this whole 4.5 acre prairie to burn. However, this year there was an area of the Pocket Prairie which burned but the tall dead stems, mostly forbs, did not burn up. This area is in the outflow of storm water events from the ravine, in a wide line between the ravine and the outflow into the culvert under Pleasant Valley Road at the south end of the Pocket Prairie. This is an area that we had recognized as wet-mesic and had planted with species suitable for that sort of soil. The unburned area shows in one of the photos.

Recap on the weather: Dry air mass came in Friday and held for weekend. Sunny. Dewpoint 21. Temp mid 30’s at night to low 50’s by noon. South wind around 10 mph.

The sun was an important factor There were no clouds during night or day as well as for the two days leading up to the burn. Because of the low dewpoint, no dew on the grass or leaves. The wind was also important in helping the fire “carry” through the savanna. The temperature is not as critical as long as above 45.

Nice comment from volunteer Ron Endres: Will and I were talking about how every single burn is different. Both the Tuesday and Sunday burns had fire behaviors we had never seen before. I think one reason we like being on these crews is that there is an element of "observe and then react" and often problem solving on the spot. There are very few "boring" fires. We also like burning with you [Amanda] and Craig and Chris. Even though you are the Burn Boss and Line Bosses, you draw on our experience. We always feel like we are needed. We work on a number of different crews and we don't always feel this way. And when the weather is like it's a wonderful day.