The Vegetation of Wisconsin: An Ordination of Plant Communities

16. Savanna

CHAPTER lG Savanna GENERAL NATURE, TYPES, AND LOCATIONS When the Spaniards began their explorations of the islands in the Caribbean, they found many types of vegetation totally unfamiliar to them. One of the most striking was a peculiar combination of grassland and forest, in which the bulk of the land was occupied by grasses and a few shrubs, but which also had widely spaced tall trees, frequently of a single species at a given place. The native Carib Indians called such landscapes "savanna." This has become the general name in many European languages for any similar plant community where trees are a component but where their density is so low that it allows grasses and other herbaceous vegetation to become the actual dominants of the community. In the tropics, savannas are largely independent of climate and are found in forest regions on soils and topographies of unusual Copyrighted Material 326 SAVANNA"'" SHRUr. CO;\[;\IlINITIES characteristics (Beard, 1953). In most of the temperate zone, however, savannas are best developed in the climatic belt separating the grasslands from the forests. Their intermediate structural nature thus parallels the intermediate climatic conditions under which they are found (Dyksterhuis, 1957). In \Visconsin, savanna constituted one of the most widespread communities in presettlement times. A number of more or less distinct types were present, which differed in the kind of trees or the kind of understory, or both. The most familiar, as well as that occupying the greatest area, was the oak opening. This savanna type was present throughout the prairie-forest floristic province, south and west of the tension zone (Figure XVI-2). The usual combination on the uplands was that of bur oak (Querclls mncrocarpa) and mesic prairie, although black oak (Q. vellililln) and white oak (Q. nliJrt) were sometimes present , while swamp white oak (Q. bic%r) and wet-mesic prairie formed the community on low grounds. A rough estimate of the area occupied by the oak openings in the early 1800's is 5,500,000 acres. Another type of sa\'anna occurred on the sand plains of central vVisconsin, on the outwash terraces of the \Visconsin and Sugar rivers, and at other places with a very sandy soil. These were locally known as "nub oak barrens. The dominant tree was Hill's oak «(( ellipsoidnlis) although true black oak was sometimes present. The understory was largely dry-mesic prairie or sand barrens grassland. Approximately 1,800,000 acres were so occupied in rresettlcment times. North of the tension zone, this type gradually gave way to jack pine (Pinus iJnnhsirll1fl) barrens, although the scrub oaks remained in the mixture throughout the northern counties (Figure XVI-4). The greatest areas of pine b:lITens were in Marinette, Oneida, Ad:lms, and Juneau counties, and in the northwestern counties in the beel of hypothetical glacial Barrens Lake (Aldrich and Fassett, 1929). The understory was either s:lI1d barrens grassland or depauperate bracken-grassland. Pine barrens origi1I :1I1y covered about 2,300,000 acres. In the south a fourth type of savanna was present. This was the very specialized cedar glade, which Oc(Upied loca I areas of steep topography , especially on limestone ledges anel clifflike slopes in the Driftless Area and on abrupt knolls of limestone drift in the Kettle Moraine and other glacial moraines. The tree member was red cedar (Juniperus virginiana ), while the understory was dry prairie. No firm estimates of the total area of this type are available, but it undoubtedly was very small, probably only a few thousand acres. The ullllsual nature of the savannas called forth almost as much Copyrighted Material discussion from early observers as did the prairie, for there were many descriptions and attempts at explanation. In spite of this, the first general summary was written as recently as 1946, by A. B. Stout, who confined his attention to the bur oak openings of southern \Visconsin. In the P.E.L. program, Cottam (1949) studied the spacing and COIll position of the oak openings in Dane and Green counties on the basis of surveyors' records. H e investigated the changes that have taken place in the openings in the past century. Brown (1950) also compared the situation when Wisconsin was originally surveyed with present cot!ditions in the jack pine barrens of Adams County. The locations of the savanna stands upon which the material in this chapter is b "0 > .. " 60 c: 40 2 o Q...