by Chris Ludwig, former Chief Biologist for Virginia Natural Heritage Program
In 1991, I bought ten acres of land in the rural Piedmont of Virginia, 35 miles west northwest of Richmond near the crossroads of Vontay in Hanover County. In the mid-1980s, the land had been selectively cut leaving some oaks and other trees so that when I bought it, the land was a jumble of stump sprouts, saplings, and brambles with scattered trees.
At the time, I was botanist with the Virginia Natural Heritage Program, and it was becoming clear to me that Virginia was losing its grassland flora. In 1992 after construction of my house, I began some pretty intense vegetation management on a couple of acres around the house. I wanted to watch how the understory vegetation might develop, and it seemed to be a wonderful way to enjoy and learn from the land around my home.
Google Earth leaf-on image of savanna shown within white polygon. Polygon measures at exactly 2 acres and is approximately 205 yards north-south. Transmission powerline right-of-way to south. House in center. Imaged May, 2018.
A lot of hard work
I started by manually lopping the woody tangle while leaving the sparse herb layer and the trees that had not been selected for harvest. These mostly ran in the 8-14”-dbh range and were a mix of oaks (white, northern red, post, southern red, willow) with lesser amounts or singles of hickory (pignut, mockernut, red), American beech, red maple, tulip-tree, black gum, sweetgum, flowering dogwood, winged elm, black walnut, American holly, eastern redcedar and Virginia pine.
June 18, 2020 – I left a few smaller trees in this corner of the savanna including young winged elm (Ulmus alata) and pignut hickory (Carya glabra) as well as a scraggly flowering dogwood (Cornus florida).
My clearing left a woodland and savanna-like appearance to the site as I supplemented the lopping with burning whenever I could do so safely and wherever the fuel load would carry fire. As a result, every square foot of the two acres has been burned at least five times and some of it has been burned as many as 20 times during the last 28 years! Because of the sparse herb layer (particularly in the early years), dry leaves were needed to carry the fire. Thus, all of the burns were conducted between December and early April.
April 18, 2020 - 3 years since burn in this shady portion of the savanna but still relatively free of woody encroachment. Light green in foreground is Japanese honeysuckle (Lonicera japonica), a persistent non-native that is tamed, but not easily eliminated, by fire.
As the savanna has developed, in addition to the lopping then burning, I have supplemented management:
· The northern one acre of the savanna has been mowed during the growing season but this was rare – some areas were mowed only three times, and no area was mowed more than six times over the last 29 years.
· Somewhat sporadically over the years, I hand pulled non-native plant species. Over the last year (as retirement has afforded me more free time) I have been much more energetic. Much of the savanna is now virtually exotic-free – I pity the non-native that has germinated in the savanna!
· Hand pulling isn’t always the answer, and I have applied Glyphosate to three species: sericea lespedeza, tall fescue, and root sprouts of sweetgum.
· As the trees have grown, shade has increased so I have also girdled a number of the less fire tolerant tree species (tulip-tree, American beech, black gum, sweetgum) that have attained a size in which they are immune to the effects of the small fires that I employ.
If this is sounds like a lot of work, it is.
Open up the Canopy
But I feel it has been worth it because I now live in a tiny Virginia Piedmont savanna and have watched the native understory develop under this management. As I have gotten to know my savanna, I have marveled at the resilience of our native grassland plants species. It is safe to say that if you open up the canopy, sun-loving herbaceous species will increase in both abundance and flowering.
But…not all species act the same! Though the ground cover is composed of predominantly perennial herbs, the composition year over year varies. Some species steadily increase or seemingly decrease. A few species seemingly hide and show up years later while a new species can pop up unexpectedly. Needless to say, after nearly 30 years, it’s a red-letter day when a new species makes itself known! Here are some examples:
April 14, 2020 – Curlyheads (Clematis ochroleuca), an uncommon perennial only found on more basic soils in this region. The species is a favorite of deer and rarely flowers at my site. Two sub-populations with just a few plants each have persisted in the same relative location of the savanna for the last 20 or more years.
July 27, 2020 – Fruiting hairy angelica (Angelica venenosa) emerging from a tangle of hog peanut (Amphicarpaea bracteata) and trailing lespedeza (Lespedeza procumbens).
· Explode and senesce: Elephantopus tomentosus, Erigeron annuus, Rhus glabra
· Barely change: Asclepias amplexicaulis, Clematis ochroleuca, Gentiana villosa, Lysmimachia lanceolata, Pycnanthemum incanum
· Steadily increase: Angelica venenosa, Desmodium viridiflorum, Elymus hystrix, E. canadensis, Leersia virginiana, Pycananthemum tenuifolium, Rudbeckia fulgida
· Seemingly decrease: Eutrochium purpureum
· Hide and show up again: Carex willdenowii, Polygala polygama
· Pop up unexpectedly (new in 2020): Calystegia spithamea ssp. spithamea, Cyperus hystricinus, Desmodium ciliare, Lespedeza bicolor (non-native since removed), Matelea sp., Physalis virginiana, Youngia japonica (non-native since removed)
July 27, 2020 - A dense patch of Black-eyed Susan (Rudbeckia hirta) in front of upland ironweed (Vernonia glauca) and Godfrey's thoroughwort (Eupatorium godfreyi).
Native to the neighborhood
It has also been instructive to watch the native graminoid cover. My experience at this site has shown me that burning does not always produce an abundance of warm-season grasses. While broomsedge, Indian grass, and little bluestem are abundant along a powerline right-of-way which is adjacent to one edge the savanna, they have been slow to colonize. In fact, only a few dozen clumps of Indian grass and little bluestem combined are found in the two-acre savanna after 28 years of burning. Even the broomsedge (which in my experience is a fast mover) has colonized less than 1% of the savanna. There are many possible explanations (no seed bank, bad timing on burns, etc.) but this trend has run counter to my hypothesis when I undertook this endeavor. While the warm season grass species have been slow to colonize, other graminoids have fared well, particularly Coleatenea anceps, Danthonia spicata, Dichanthelium boscii, Elymus hystrix and E. virginicus, Leersia virginica, Scleria oligantha, Sphenopholis nitida, S. obtusata, and Tridens flavus.
Sept. 17, 2020 - although we have had plenty of rain in 2020 and not yet had a frost, the vegetation is less vigorous and browning this time of year. Many early season species can no longer be found. The light brown in the upper left is Indian grass (Sorghastrum nutans) vigorously flowering/fruiting in the adjacent transmission powerline. Curiously, Indian grass is limited in the savanna. In the distance near the house, tall goldenrod (Solidago altisimma) not yet in full flower forms a light-yellow haze above the richer yellow of peak blooming gray goldenrod (Solidago nemoralis).
If you have some native grassland species to start with, I strongly suggest NOT adding native plants or seed. It takes more patience, but you can learn a great deal by watching the native flora express itself over the years. If after a year or two no natives express themselves, then adding seed plants may be your best choice.
Many was the time, particularly in the early part of this work, when I craved a more vibrant and lusher display of grassland flora, and I totally understand why many would plant natives to jump start their site. For me, however, the wait has been worth it, because (as you can see from the pictures) I now have a fantastic display of native plants that aren’t only native to my state or ecoregion, they are actually native to the neighborhood!
I could write volumes about the non-native plants that I am striving to eliminate, but I have learned two lessons worth sharing. I am sure you have heard the first before: Never put off ‘til tomorrow what you can do today. Too many times over the years of watching the savanna have I noticed a non-native and said, ‘I will get that next time.’ Well when next time rolls around, you may find 10x or 100x the size of the problem you ignored in the first place. At my place, Anthoxanthum odoratum, Arthraxon hispidus var. hispidus, Commelina communis, Lespedeza cuneata, Kummerowia striata, Microstegium vimineum, and Trifolium repens fall into this category.
Get them now!
My second offering on the non-native front: Pulling weeds can be made fun! I have found more than one new native by carefully looking for and pulling non-natives. It slows you down. The other fun comes when you slow your mind and enjoy the Zen aspects of pulling. And of course, there is the reward of looking out over a sea of native plants!
Sept. 29, 2020 - 9 species of Tick-trefoil (Desmodium/Hylodesmum) fill the late summer understory of the savanna. Wonder why?
256 natives on 2 acres
In preparation for writing this, I used 2020 to concentrate and inspect every square inch of the two acres – developing a complete list of the vascular plants (while mercilessly ripping up any non-natives!) Here I provide a complete list of the vascular plants I could find in 2020 with frequent puttering through the tangle. I topped out at 301 vascular plant species. For grins, I kept track of month when most of the flowering plants started blooming.
Of the 301 species, 45 were non-native and my goal is to get this number to zero. If you’re interested, check back in with me in a few years – I will let you know of my progress.
That leaves 256 native species - a fairly hefty number of vascular plant species for two Virginia acres. As I puttered through the year, I wondered if Virginia’s Piedmont pre-settlement burned savannas might match this number on any random two acres. I believe the answer would be yes if the random patch had three characteristics found in my savanna.
First, the savanna would need to be on basic soil. At home, my soils are derived from basic igneous rocks of the Virginia Piedmont including amphibolite. While acidic soils might capture a few heath, oak, or sedge species lacking from the species list for this site, basic soils are favored by any number of species found here including Clematis ochroleuca, Gentiana villosa, Liatris squarrosa, and Triosteum angustifolium. While legume diversity is generally higher in savanna habitats in Virginia, legume species counts are highest over basic soils.
Second, 256 native species in a random two-acre patch of savanna is more likely when a mechanical disturbance regime is in place or nearby. My site is adjacent to our yard, a gravel driveway, and a transmission powerline right-of-way. Here, a subset of the native savanna species was observed to be more prevalent near the more frequently disturbed habitat along these edges. Native plant species in this category include Ambrosia artemisifolia, Euphorbia maculata, Oenothera biennis, Paspalum pubiflorum, Phyllanthus caroliniensis, and Verbesina alternifolia. On our precolonial landscape, savannas with mechanical disturbance would include areas adjacent to herbivore migration trails as well as Indian settlements, croplands, and transportation corridors.
July 16, 2020 - Carolina Wild-petunia (Ruellia caroliniensis) that is slowly coming in to the savanna from the adjacent powerline right-of-way.
Third, my savanna also captures a few forest herbs that are found in adjacent shaded habitats including Athyrium asplenioides, Liparis lilifolia, Polystichum acrostichoides, Smilax pulverulenta, and Tipularia discolor.
Imagining the diversity of precolonial Virginia savannas and enjoying one now
In further support of a diverse precolonial savanna: my two-acre patch hardly catches all the potential species that are present in similar patches nearby. On the transmission powerline and a bank along a gravel road less than 100 yards from the savanna, I have watched many sun-loving species for 28 years fail to colonize in the seemingly suitable habitat I have created in my savanna. These include Agalinis purpurea, Aristida longespica, Ascelpias viridiflora, Crotolaria sagittalis, Gymnopogon ambiguus, Helianthus atrorubens, Heliathus divaricatus, Heliopsis helianthoides, and Pyrrophappus carolinianus. A ride on the backroads within a few miles of the house reveals other savanna species not found at my house including Andropogon ternarius, Asclepias syriaca, Asclepias purpurascens, Carex meadii, Coreopsis auriculata, Coreopsis verticillata, Helianthus giganteus, Liatris pilosa, Linum intercursum, Lobelia puberula, Parthenium integrifolium, Paspalum floridanum, Scleria pauciflora, and Tephrosia virginiana.
July 27, 2020 – Blooming scaley blazing star (Liatris squarrosa) below with upland ironweed (Vernonia glauca) blooming above.
The point is that a large reservoir of potential native vascular plant species exists that any random two-acre savanna may have captured in the precolonial Virginia Piedmont, particularly if soils were basic and that area was adjacent to forests and mechanical disturbance. Could there have been a 300- or 350-species two-acre savanna in the Piedmont of Virginia? Perhaps so. What if the two acres had basic soils, adjacent disturbance and forest, and intersected a wetland?
If you love grasslands as much as I do and want to learn more about oak savanna restoration, I highly recommend this recent paper which quantifies and expands on many of my ramblings above, by Vander Yacht et al. The title alone, “Litter to glitter: promoting herbaceous groundcover and diversity in mid-southern USA oak forests using canopy disturbance and fire,” stirs one’s heart.
May 22, 2020 - Under the shade of a post oak (Quercus stellata).
April 14, 2020 - it's not just about the plants! Sleeping eastern tailed blue butterflies (Cupido comyntas) 8:45am. This butterfly species is quite common in Virginia and this savanna and is often referred to as an “ETB”. ETBs aren’t particular about their nectar sources but they only use pea family (Fabaceae) as their host plants. Thirty species of pea were found at this site in 2020.
In addition to watching over the plants, many other joys have presented themselves over the years in my grassland, including watching a Woodcock “peent” in the black of an area burned in early spring, observing Turkey and Bobwhite regularly visit (though the Bobwhites have thinned out sadly), and finding the only Virginia record of a species of dagger moth, Acronicta albarufa. This globally rare noctuid moth is believed to be a post oak obligate. The butterflies and October flight of buck moths (Hemileuca maia) are a joy as well.
I would encourage any lover of grasslands lucky enough to have a plot to give this a try. It is labor-intensive but so very rewarding if you love our native grasslands and would like to have one to study from your back door!
July 22, 2020 – About half of the savanna is shown in this photo. Fern in lower left is Christmas fern (Polystichum acrostichoides).