Many of todays oak forests are under stress from various factors including defoliation by gypsy moth, oak dieback and decline, and possibly global warming. All of these stresses influence and potentially limit the flexibility we have in sustaining and managing oak forests. The various threats to oak forest health are reviewed in the following sections:

Acorn-infesting insects are the most studied group of pests affecting oak regeneration. Weevils of the genus Curculio are the major culprit (Barrett 1931; Beck 1977; Beck and Olson 1968; Collins 1961; Downs and McQuilken 1944; Kearby and others 1986). However, Conotrachelus weevils, cynipid gall wasps, and both primary and secondary Lepidoptera have all been implicated (Kearby and others 1986). Infestation rates vary considerably from year to year, with nearly 100 percent in some years. In several long-term studies, infestation by insects averaged around 50 percent, a very significant impact (Beck 1977; Burns and others 1954; Christisen 1955; Downs and McQuilken 1944; Kearby and others 1986; Beck, 1993).

Virtually all oak species are attacked by one or more of 22 different weevils recorded by Williams (1989). Adult weevils are themselves not damaging. However, larvae hatching from eggs laid in late summer in tiny niches beneath the shell may consume most of the nut tissue within a few weeks. When larvae mature, they bore out through the shell and migrate underground to pupate. The rate of infestation is variable but has exceeded 90 percent in some northern red oak collections (Gibson 1982). Single acorns are commonly hosts to three or more larvae; embryos in infested acorns that escape damage may germinate but seedlings grow slower than those from uninfested acorns (Oliver and Chapin 1984; Oak, 1993).

Infestation rates of the filbertworm (Melissopus latiferreanus) are much lower than for acorn weevils (Gibson 1982), but they have been responsible for large losses, particularly in low production years (Drooz 1985). Damage is caused by larval feeding and is usually lethal to infested acorns. Adult moths lay single eggs on leaf surfaces near acorn clusters in mid-summer. After hatching, the larva crawls to an acorn, bores through the shell, and begins feeding. The mature larva exits the acorn after 3 to 5 weeks, spins a cocoon in the top few inches of soil, and pupates in winter (Oak, 1993).

Non-stinging gall wasps are another group of primary acorn pests. Like Curculio spp. and the filbertworm, they infest and can kill intact acorns. There are many species that form unusual galls on various oak tissues, but Callyrhytis operator and C. fructuosa are the most common in acorns. Gibson (1982) found them to be less ubiquitous than either weevils or filbertworms in northern red oak acorn collections. However, the infestation rate of C. fructuosa was second only to weevils in two collections made in consecutive years from a Tennessee seed orchard (L. Barber and author, unpublished data1). Acorns infested by C. fructuosa appear normal on the outside but are filled with up to 2 dozen larvae encased in small stony galls. C. operator, on the other hand, forms a gall in the side of the acorn shell. Gall wasps have complicated life cycles, and the same species may induce galls on different plant parts at different times of the year (Oak, 1993). In a Missouri study the primary moth invader, M. latiferreanus, and cynipid gall wasps infested 17 percent of immature acorns collected in a 4-year study (Kearby and others 1986; Beck, 1993).

Survival Curve of the Pistillate Flower

Some insects and pathogenic fungi and bacteria can invade acorns damaged by other agents and thereby increase losses. The best known insects with this mode of action are Conotrachelus weevils and the acorn moth, Valentinia glandulella. Neither can breech intact acorns, but the acorn moth has been known to attack otherwise healthy, germinating acorns (Galford 1986). Examples of pathogens following acorn damage are a bacterium, Erwinia quercina, and a fungus, Fusarium solani. E. quercina causes a disease in California live oak (Q. agrifolia), called drippy nut, after it gains access to acorns through oviposition punctures of gall wasps (Hildebrand and Schroth 1967). The name of the disease describes the main symptom, which is the byproduct of anaerobic fermentation and results in acorn rot and premature abscission. F. solani was isolated from weeviled acorns collected in Mississippi (Vozzo 1983) and Tennessee (author, unpublished data2). The fungus can cause damping-off and root disease in seedlings, but its impact in these cases is unknown since saprophytic forms exist and pathogenicity was not confirmed experimentally (Oak, 1993).

Most accounts of acorn consumption by insects are about weevils (Curculio, spp. and Conotrachelus spp.) as they destroy acorn crops (Gibson 1964, Kearby et al 1986). We observed these insects feeding on flowers that then died 1 week later (Cecich et al 1991). Controlled feeding experiments are being done at this time to better elucidate the role of treehoppers in flower abortion (Cecich, 1994).

Mature acorns that escape destruction by insects face yet another hurdle. Nearly 200 species of forest wildlife consume acorns. Some, such as the large game animals, consume large quantities as individuals and in aggregate. But the sciurids (chipmunks, squirrels, etc.) probably have the largest impact (Johnson and others 1989). In looking at cumulative mast needs by forest wildlife in the southern Appalachians, they said that acorn crops of <200 lbs/acre usually are totally consumed. And, over a period of years the demand will always exceed the supply. In a 12-year study in the southern Appalachians at one location, production of sound acorns exceeded 200 lbs/acre only four times (Beck 1977). Average production was 186 lbs/acre and that number was influenced heavily by one bumper year with 800 lbs produced. Even in poor mast years some acorns escape predators and produce seedlings. But it takes a bumper-crop year for any appreciable number to do so (Beck, 1993).

However, an acorn that disappears from a seed trap should not necessarily be considered "destroyed" (Lorimer, 1993). Thorn and Tzilkowski (1991) point out that by burying acorns in well-distributed caches, small mammals may actually facilitate seedling germination. In that study, small mammal activity resulted in 28,000 well-distributed caches per acre.

Tree seedlings grown in large numbers in confined areas under the lush growing conditions typical of nurseries are subject to damage from a whole suite of pests not normally found in forested settings. Defoliating insects, foliage disease, shoot cankers, and root diseases can cause dramatic symptoms in nurseries. However, the most widespread and damaging of oak nursery pests are root diseases caused by fungi in the genera Phytophthora and Cylindrocladium. They have tough resting spore stages that can survive very harsh conditions and persist in soil for long periods in the absence of suitable hosts. P. cinnamomi is a virulent pathogen of many oak species in nurseries, most notably northern red oak (Crandell and others 1945). It is a member of a group of fungi known as water molds that are favored by periodically wet soil conditions and have a swimming spore stage that attacks feeder roots. Once established in the root system, the fungus spreads to larger laterals and the tap root. Cylindrocladium spp. have no mobile spore stage but they form densely compacted masses of fungal tissue in soil and colonized plant parts called sclerotia that resist all but the most aggressive of control measures. Several different Cylindrocladium spp. have caused root rot in cherrybark (Smyly and others 1977) and shumard oaks (Affeltranger and Burns 1983) in bareroot nurseries and in northern red oak grown in containers (Oak and Triplett 1985; Oak, 1993).

Root-diseased seedlings may display a range of above-ground symptoms including pre- and post-emergence damping-off, stunting, top dieback, foliage yellowing, and premature defoliation. Below ground, infected root systems are sparse, stunted, and usually blackened. Obviously diseased seedlings are easily recognized and culled, but those with less distinct symptoms may be outplanted and die later or grow slower than healthy seedlings (Oak, 1993).

Control of Nursery Root Diseases

Nursery root disease losses have been most effectively avoided where sanitation, soil management, and chemical measures have been combined into an integrated pest management system (Cordell and others 1989). Sanitation starts with excluding the pathogens that can be transported to uninfested areas in contaminated soil by thoroughly cleaning equipment that has been used in infested areas. If pathogens become established, then a vigilant inspection schedule, culling of affected stock, and in extreme cases, quarantine, can reduce losses. Soil management practices that have an important influence on root disease probability include crop rotation, choice of cover crops, managing water, and in containerized nurseries, choice of growing medium. Continuous seedling cropping leads to increased populations of pathogens, and some cover crops are also hosts for seedling pathogens. For instance, legume crops increase inoculum levels of Cylindrocladium spp., but grass cover crops do not (Soloman and others 1987). Inadequate drainage or over-watering can simultaneously increase inoculum of Phytophthora spp. and increase seedling susceptibility by depressing seedling vigor and inhibiting normal root development. Commercially prepared "artificial" growing media are preferred in container systems over mixtures that include field soil, due to the risk of introducing pathogens and the difficulty in eradicating them once introduced (Oak and Triplett 1985; Oak, 1993).

Root pathogens are among the most difficult nursery pests to control chemically. Several soil fumigants are EPA-registered but the most effective are formulations containing 67 percent methyl bromide and 33 percent chloropicrin (Cordell and others 1989). Proper soil preparation, soil moisture, temperature, and deep placement of the fumigant are essential for effective treatment. Fungicide drenches applied to the soil (e.g. metalaxyl for Phytophthora spp. and benomyl for Cylindrocladium spp.) are registered for nursery sites and are widely prescribed, but the research basis for recommending them to control root diseases in oak seedlings is lacking (Oak, 1993).

A malady of oaks called oak decline is a widespread and a potentially serious threat to millions of acres of oak forests.  Its effects range from partial crown dieback to tree death.  This phenomenon is believed to be initiated by drought stress and/or defoliation, which is subsequently exacerbated by further insect and disease attack (Wargo and Haack, 1991).  Maintaining species diversity is believed to be important in minimizing the impact of this stress-mediated disease (Houston, 1987).  The most susceptible stands are those of advanced age occurring on drier sites (red oak site index < 70 ft), and largely comprised of trees in the red oak group (Starkey et al., 1989).  Because tree age and site quality are the most important factors that predispose oaks to decline, the ratio of site index to tree or stand age is a useful index of their susceptibility to decline; low ratios are associated with a high susceptibility (Oak and Starkey, 1991){Johnson 2002}. 

Although much remains to be learned about oak decline, possible preventive or remedial silvicultural treatments include: (1) thinning early to reduce future competition-induced stress and to favor less susceptible species, (2) reducing rotation age, (3) favoring less susceptible species when regenerating stands, and (4) controlling defoliating insects (Starkey et al., 1989).  There is some evidence that thinning stands where decline is already present increases its severity.  Possible explanations include an increased food base for Armillaria root disease associated with the root systems of harvested and logging-injured trees, reduced soil moisture related to soil compaction from logging, and increased temperatures on the forest floor related to reduced overstory density (Starkey et al., 1989).  Thinning nevertheless reduced the incidence of oak decline in a Missouri study (Dwyer et al., 1995).  Although thinning should theoretically increase the vigor of residual tees, they may be temporarily stressed by their sudden exposure to increased sunlight, wind, raised water tables, winter injury, and increased soil temperatures.  In turn, these stresses may reduce their resistance to pathogens such as Armillaria root disease before the benefits of thinning occur (Wargo and Harrington, 1991).

Oak decline, whose effects on oaks range from partial crown dieback to tree death, also afflicts northern red oak. This phenomenon is believed to be initiated by drought stress and/or defoliation, which is subsequently exacerbated by further insect and disease attack (Wargo and Haack 1991). Maintaining species diversity is believed to be important in minimizing the impact of this stress-mediated disease (Houston 1987). Stands at greatest risk are older stands containing a large proportion of trees in the red oak group (subgenus Erythrobalanus) that occur on xeric sites (red oak site index [base age 50] <70) (Starkey et al 1989). Because tree age and site quality are the most important factors that predispose oaks to decline, the ratio of site index to tree or stand age is a useful index of their vulnerability to decline; low ratios are associated with a high incidence of decline (Oak and Starkey 1991). (Johnson, 1994) {Johnson 2002}

Most references to oak regeneration problems are for high-quality sites, but oak decline has the potential to limit natural regeneration opportunities in mature oak stands growing on drier sites. There is no doubt that large areas are being affected by oak decline. Over 1 million acres of oak forest type were affected in the northern Piedmont and Appalachian Mountains of Virginia in 1986 (Oak and others 1991). Preliminary estimates for the Piedmont and mountains of North Carolina were about .9 million acres in 1990 (R. Sheffield, unpublished data). The regeneration impacts are less clear, however. Acorn production potential is reduced as a result of crown dieback and tree mortality (Oak and others 1989). Advanced physiological age, root disease, and carbohydrate physiology prevent vigorous sprouting of overstory trees in decline areas. Whether these effects result in an inability to adequately regenerate oak depends on the abundance and competitiveness of oak propagules relative to other species that might displace it. Monitoring of a declining scarlet oak/black oak stand in Missouri over 5 years showed overstory mortality increasing from 36 to 62 percent and shifts in understory composition towards white and post oaks, red maple, and shortleaf pine (Johnson and Law 1989). Oaks will probably be a smaller component in the next stand and oak species diversity will be less. Similar conclusions were reached after 8 years of monitoring in gypsy moth defoliated areas of the Pocono Mountains in Pennsylvania (Gansner and others 1983), where the greatest oak losses were in areas suffering repeated, heavy defoliations. Pre-defoliation decline conditions were not reported. On the George Washington National Forest in Virginia, the combined effects of pre-existing oak decline and gypsy moth defoliation have resulted in 22,500 acres of severe mortality accumulated between 1987-1991. Research is needed into the interactions of oak decline, insect defoliation, overstory competition, understory composition, and regeneration timing as well as options for dealing with stands suffering severe mortality from decline and defoliation (Oak, 1993).

A silvicultural strategy was developed for mixed oak stands in the Ozark Highlands that are at risk for oak decline (Dwyer and Kurtz, 1994).   Stands dominated by scarlet and black oaks of advanced age, which are common in this region, are especially susceptible to oak decline (Jenkins and Pallardy,1995; Wetteroff, 1993).  These stands occur on sites ranging from oak site index 60 to 70 ft (McQuilkin, 1974).  The strategy calls for  removing oaks with 30 percent or greater crown dieback as a first priority measure, and secondly, the removal of smaller overtopped scarlet and black oaks.  Thinning is deferred to stand age 40 at which time stand density is reduced to 65 percent stocking based on Gingrich’s (1967) stocking chart (Fig. 6.9).  This thinning is expected to removed about 450 board feet of sawtimber and 6 cords of pulpwood per acre.  A second thinning is made at stand age 50 by reducing stocking to 64 percent, which removes about 950 board feet of sawtimber and 2 cords of pulpwood per acre.  At stand age 60, a shelterwood harvest that reduces stocking to 44 percent is recommended.  The final harvest is at age 63, which is expected to remove about 3,300 board feet of sawtimber and 4 cords of pulpwood per acre.  The overall strategy is based on comparing various thinning alternatives and is designed to maximize the net present worth of a stand after taking into consideration the risk of decline-related mortality, returns from thinning, the carrying costs of thinning, and final crop tree value. 

Both clearcutting and shelterwood methods are potentially useful for regenerating oak  stands of advanced age that are decline-susceptible or that are already affected by decline.  Clearcutting may be the more appropriate choice in stands largely comprised of red oaks and where the current stand regeneration potential is sufficient to meet management objectives, or supplemental oak planting is feasible.  In stands already affected by oak decline, the stump sprouting potential of the overstory has already been reduced.  Expected contributions to future stocking from oak stump sprouts therefore need to be proportionately reduced to account for  the number of decline-affected trees. {Johnson 2002}

Global warming and associated summer drought and winter freeze-thaw cycles also have been implicated in the increased occurrence of dieback and decline of forest trees across North America (Auclair et al 1990). Although oaks are known for their great drought tolerance, recent research has shown that, to the contrary, they are vulnerable to sudden and fatal disruptions in xylem water conduction (embolisms) induced by water stress, especially stress caused by winter desiccation (Cochard and Tyree 1990). Those effects eventually may fit into a more comprehensive and fundamental decline theory than those currently proposed (c4: Ammon et al 1992; Auclair et al 1992; Houston 1992; Manion and Lachance 1992; Mueller-Dombois 1992; Wargo and Haack 1991; Johnson, 1994).

Although much remains to be learned about oak decline, possible preventive or remedial silvicultural treatments include:

1. Thinning early to reduce future competition-induced stress and to favor less susceptible species, 
2. Reducing rotation age, 
3. Favoring less susceptible species when regenerating stands, and 
4. Controlling defoliating insects (Starkey et al 1989). 

There is some evidence that making partial cuttings in stands where decline is already present increase the severity of decline. Possible explanations include an increased food supply for shoestring root rot originating from root systems of stumps and trees injured during logging, reduced soil moisture related to soil compaction from logging, and increased temperatures on the forest floor related to reduced overstory density (Starkey et al 1989). Because highest mortality of oaks from decline occurs on droughty sites, populations of northern red oak may be affected less than the more xerophytic oaks such as black oak and scarlet oak. Nevertheless, a high incidence of decline in northern red oak has been observed in some areas of the southern Appalachians (Tainter et al. 1984; Johnson, 1994).

Oaks are the preferred host of the gypsy moth, and thus stands most susceptible to attack are those with a high proportion of oaks (Gottschalk 1991). Northern red oak ranks behind chestnut oak (Quercus prinus L.), black oak, and scarlet oak as a preferred host (Gansner and Herrick 1985). In all species, heavy defoliation may cause trees to refoliate during the same year, which in turn may deplete carbohydrate reserves and lead to attack by secondary organisms such as shoestring root rot (Armillaria spp.) and twolined chestnut borer (Agrilus bilineatus) (Gansner and Herrick 1985, Gottschalk et al 1989). Trees at highest risk are those in suppressed (overtopped) and intermediate crown classes and, within a given crown class, trees with low-vigor crowns (Herrick and Gansner 1987). Mortality associated with gypsy moth defoliation also is greater on good sites than on poor sites (Gansner 1987), and under some conditions is greater in thinned stands than in unthinned stands (Twery and Gottschalk 1989; Johnson, 1994) {Johnson 2002}.

The most significant effect of the gypsy moth on the oak resource is not mortality, but reduced growth, yield, and wood quality. Gypsy moth outbreaks occur at 7- to 10-year intervals and last 1 to 2 years (Gottschalk et al 1989). It may take 3 years for stand growth to recover from a single defoliation (Twery 1987). For two defoliations in 1 decade, loss of volume increment has been estimated at about 10 percent, exclusive of mortality. For two defoliations per 5 years, estimated growth losses are about 19 percent (Twery 1987). However, such losses may be offset by subsequent gains in growth of the survivors related to the thinning effect from mortality. To minimize loss in value of sawtimber-size oaks killed by defoliation, trees should be salvaged within the first year after death (Garges et al 1984; Johnson, 1994).

Strategies for reducing timber losses from gypsy moth include reducing the proportion of oaks in the stand, applying insecticides, and removing trees that are preferred hosts and refuges for the insect. The latter include oaks and other species in poor crown condition. Specific strategies and options for coping with the gypsy moth are varied and depend on stand condition and age, the current status of the insect in the stand, and management objectives (Gottschalk 1987. 1988, 1989, 1991, 1993; Johnson, 1994).

The presence of gypsy moths will have a long-term effect on the silvicultural practices used in eastern hardwood forest management. However, some natural resistance to defoliation will occur and, though the oak composition in Appalachian forest stands will be significantly reduced, it is not expected to be completely eliminated by the moths. Many oak trees still remain associated with the gypsy moths. The forest managers inability to reproduce oak on these good growing sites may have a negative impact on moth populations. The stands will become more mixed with other hardwoods, thus minimizing the attractiveness to the moths. Stands on forests with fewer oak should not be able to sustain the moths at frequent, epidemic levels (Steve Jones, personal comments.) (C. Smith, 1993).

Current information about Gypsy Moth can be found at
http://www.fs.fed.us/ne/morgantown/4557/gmoth

Deer populations in certain areas of the central Appalachians are so high, that understory browse lines are quite visible. As the deer population increases, there will be public, and thus political, resistance to increasing deer control measures for the sake of forestry purposes. In the future, much of the Appalachian forests could have a browse line similar to parts of the northeastern forests. When this occurs, it is too late for quick remedial actions. (C. Smith, 1993)

The potential for deer browsing to block the development of competitive oak advance regeneration is better documented than insect damage. The deer population on the Allegheny Plateau in Pennsylvania (often more than 30 per square mile) is sufficient to create open, park-like stands with little undergrowth. Where deer populations are high, browsing can occur on oak seedlings that are as short as 15 in. high (Galford and others 1991). Sometimes there is so little advance regeneration of hardwoods that clearcuts revert to grass and scattered shrubs (Marquis 1974). (Lorimer, 1993)

Published evidence on the effects of deer browsing on oak is very limited outside of Pennsylvania. Similarly severe effects have been documented elsewhere, but the problem is often quite localized. High deer populations (34-59 per square mile) in a game preserve in Massachusetts have created savanna-like conditions, but in the surrounding region the deer average only 3-8 per square mile and browsing is limited (Healy and Lyons 1987). The intensity of deer browsing appears to vary greatly from place to place. Four underplanting trials of northern red oak in southern Wisconsin have shown little browsing in two counties with average deer populations of 18 per square mile (Pubanz and Lorimer 1992), but destructive levels of browsing in counties with average deer densities of 25-35 per square mile (Pubanz and Lorimer, personal observations). In the mountains of West Virginia, moderate-sized clearcuts (e.g., 20 acres) develop so much vegetation that the ability of deer to modify the outcome is limited (H. C. Smith, personal communication). While deer browsing was observed on oak seedlings in a southern Appalachian cove stand, and may have contributed to the slow growth rate (Beck 1970), the problem of slow growth persisted long after the deer density had greatly diminished (D. E. Beck, personal communication). (Lorimer, 1993)

Deer browsing is clearly a limiting factor for oak regeneration in some places, and the substantial growth of deer populations that occurred in many areas around the 1930s does coincide with the beginning of widespread oak problems. However, the occurrence of oak regeneration failures in places where deer are not especially numerous makes a number of researchers feel that deer are generally more of an aggravating factor than a primary limiting factor. We need more evidence, however, on the effects of moderate deer browsing on growth rates, especially where deer may be browsing oak in preference to other species (George and others 1991). Furthermore, the deer problem seems to be getting progressively worse. Deer populations in the lower Midwest were historically low (USDA Forest Service 1970) but appear to have increased substantially in recent years (e.g., Ishmael 1990). As one forester in Indiana remarked, complaints about deer damage are becoming more numerous, and establishing a hardwood plantation is often like "setting the table for deer." (Lorimer, 1993)

Although large numbers of oak seedlings may become established after bumper acorn crops, mortality rates of young seedlings are high, especially on mesic sites (Johnson 1985). A newly established cohort of seedlings in a mature stand on a good site in the Southern Appalachians had a 10-year survival rate of 10 percent, with negligible growth of the survivors (Loftis 1988a). Could repeated defoliation and other animal damage be a factor in causing high mortality and slow growth? (Lorimer, 1993)? Relatively little evidence is available on light ambient levels and resulting impacts of insect defoliation of oak seedlings or deer browsing over large areas. (Lorimer, 1993)

Evidence on seedling damage by insects is very fragmentary. An insect with a potentially serious impact on oak seedlings is the Asiatic oak weevil (Cyrtepistomus castaneus) because larvae feed on fine root hairs and adults feed on leaves (Triplehorn 1955, Roling 1979). The weevil, introduced from Japan and first recorded in New Jersey in 1933, is now distributed in most of the eastern states as far west as Missouri and Kansas. In Missouri, the weevil has a strong preference for oak and apparently has little impact on other hardwoods (Ferguson 1987). Linit and others (1986) reported 46 insect species, including the Asiatic oak weevil, associated with planted oak seedlings in Missouri. Leaf area losses averaged about 22 percent over a season. The impact on growth was not measured, but was not necessarily considered to be severe. (Lorimer, 1993)

In addition to obtaining more evidence on the impact of low to moderate defoliation on growth and vigor, more evidence is needed on insect activity across habitat types. The introduction of the Asiatic oak weevil in about 1930 did coincide with the start of major oak regeneration failures, but for that to be considered a principal limiting factor, evidence would probably be needed that weevil activity is concentrated on mesic sites where most oak regeneration failures occur. (Lorimer, 1993)

Oaks are usually considered to be favored by relatively warm and dry climates, and paleocologists often use abundant fossil oak pollen as one indicator of this type of climate. Oak regeneration failures could in principle be related to a shift to cooler and moister conditions. This change would probably have an effect of a southward shift in the boundary between northern (beech-maple) and central hardwoods (oak-hickory). In other regions, it might also cause an expansion of mesophytic species onto sites that were once relatively warm and dry, with a corresponding shrinkage of microhabitats that were formerly stable oak sites. (Lorimer, 1993)

The issue of possible climatic effects needs further study, but several important lines of evidence seem rather inconsistent with the hypothesis that climatic change has been a major causal factor. There is little evidence in the pollen record that a rather dramatic cooling episode known as the "Little Ice Age" (ca. 1400-1850) had any substantial impact on oak dominance (Davis 1958, Watts 1979, Patterson and Backman 1988). Indeed, most early explorers in the Northeast described oak as the dominant forest type. Second, data show that the climate has been warmer (and drier) in recent decades compared to the late 19th and early 20th centuries (Wahl 1968), a trend that should not be favorable for the shade-tolerant competitors. Third, a relatively warm and dry climate does not seem to facilitate oak regeneration on mesic sites or hinder the development of mesic competitors to any noticeable degree. This relationship is apparent from the development of dense understories of sugar maple on mesic sites in southern Illinois, southwestern Wisconsin, and the River Hills section of Missouri, all of which were at or close to the original prairie border (Schlesinger 1989, Pallardy and others 1988, Hix and Lorimer 1991). (Lorimer, 1993)