How did you get interested in birds?

I had to do a book report in the fourth grade, so I picked out the skinniest book I could find in the library, the Golden Field Guide to Birds. Around the same time, our local dentist was giving out little cards featuring different kinds of birds for good behavior. It all came together when I was walking home from school one day in that same year. It was a warm day-in Michigan in late winter, that's hard to come by-and cedar waxwings were feeding in the multiflora rose bushes in the park. I knew what they were, from having done that little bit with the bird book. That was 1959, and I have been watching birds ever since.

You seem to have gone from books to nature, rather than vice versa . . .

I went to a residential high school in upstate New York. During my sophomore year, I found a book in the library called An Introduction to Ornithology written by George Wallace, the ornithologist at Michigan State University. They let me check it out for the summer, and I actually read it all over that summer. My senior year I did a breeding bird census project on the grounds of the school and got those data published. It was really fun.

When I started college at Michigan State, there wasn't much doubt in my mind that I was going to be a biologist, but I didn't exactly know what I was going to do. I latched on to different professors, but my curiosity always went towards birds. When I graduated, I had a choice between working with that same professor, George Wallace, or the herpetologist. Southern Michigan is a better place for somebody who studies birds than for somebody who studies salamanders, so I ended up working on birds.

What led you towards your current work?

In 1984, I got a job as the zoologist for the Tennessee Natural Heritage Program maintaining a database of records of rare animals throughout the State. We looked at where these animals might be, the habitat they might use, the kind of environmental features that might indicate where to find them. We were interested in how we could gather those data quickly, and then translate them into useable information for environmental protection. Part of the job was verifying records. I have a predilection for survey work, and I got involved in cooperative projects with other agencies doing what we called natural area surveys, looking in different parts of the State to identify areas valuable for rare species protection that we didn't already recognize or about which we could improve our knowledge.

How did you get into cerulean warbler research with the Forest Service?

In February 1993, Winston Smith hired me to work for the Forest Service at the same place where I am right now-Stoneville, MS. The cerulean warbler project was already in existence before the job came open, and I had been cooperating on it while I was in Tennessee. I had encountered cerulean warblers, in small numbers, in the surveys we did in both middle and west Tennessee. Two people involved in a conference in 1989 in Massachusetts on migratory birds, Chan Robbins and John Fitzpatrick, had started writing a paper on cerulean warbler because they recognized that the birds were in trouble, and that we didn't know very much about them. I had some data from our work in Tennessee that I could add, so we wrote a paper that was published in 1992.

What inspires you about the research you do?

I don't know, except that it's really fascinating. What immediately popped into my mind when you asked the question is that it seems like Murphy (of Murphy's law) is in charge. In any extended field project, what we thought we understood at the end of the last iteration of activity in some respect no longer applies in the next one. It's fascinating to keep going, and try to understand what may have been underneath the change in perception that occurred. What was it that led to the assumption I had last time that now needs to be changed in light of what I have just observed? In some situations, it just takes time to amass enough information to be able credibly to address questions that may seem so obvious in the early stages that they don't get evaluated carefully.

Can you give me an example?

When I was in Tennessee, I assumed that cerulean warblers had to be in woods that were pristine and untouched, so old-growth forest would be the best place to look for them. And because I believed that to be the case, I was convinced that any sort of forest management would have a negative effect on their environment. That's just wrong. I didn't have enough experience, and I didn't know enough about silviculture. It's been a rich education to be the wildlife biologist in a silviculture unit, where I'm surrounded by people who understand the biology and economics of trees-and to learn things that as a city boy and as an environmental protection specialist I missed out on in terms of the dynamics of forests. Through working with some very smart people in different areas of expertise, I've realized that it's just not true that forest management is bad for cerulean warblers, and in fact, forest management has an important role to play in maintaining habitat on the landscape.

What is that role and how does it play out?

It's really about the interaction between silviculturists and bird biologists. What silviculturists bring is a knowledge of forest growth patterns and ability to manipulate stands of trees from a current condition in terms of species composition and vegetation structure to a more desirable condition that meets specific management objectives, for example, habitat for cerulean warbler. What the bird and salamander people and those interested in other inhabitants of the forest bring is information about how to specify what that desired condition is. Once it's defined, there's the opportunity to work with the silviculturist to devise ways to groom the forest from the current condition to the desired condition. Now, because those of us who are interested in birds have learned something about silviculture, we can have that interaction. I think it's been beneficial in both directions. The role of forest management with respect to cerulean warbler is to maintain the diverse physical vegetation structure that is preferred by the birds.

What's your favorite project?

Whatever I'm doing today ends up being my favorite project. It really is true: I'm fascinated by the work on small mammals while we're doing that, and I'm fascinated by the work on winter bird populations while we're doing that. I'm equally fascinated by the work on rusty blackbirds, cerulean warbler, and on pondberry. It's all good.

What gives you hope?

Two things: One is the interactions possible among diverse groups of people. The Cerulean Warbler Technical Group is an absolutely wonderful example of people working together, with interaction from forest industry, from nongovernmental organizations, from the coal mining industry, from numerous State and Federal agencies, and from universities in North America and South America. That's really hopeful.

The other is the really intense and high-quality interest of Latin Americans in bird study. I see the quality of the science that's being done by colleagues across South America. There are wonderful scientists, young scientists, in these countries, and that's exciting as we watch human population go up and the proportion of landscape in forest go down. It's not an entirely hopeless situation. There is a future.

In the Daniel Boone National Forest near Cold Hill, KY, a team of Forest Service (FS) and university researchers and national forest technicians have started an ambitious experiment to find out whether manipulating the structure of forest stands before the gypsy moth arrives can reduce damage from the pest. Set up as a silvicultural assessment under Title IV of the Healthy Forests Restoration Act (HFRA) of 2003, the project will test the ability of four different silvicultural treatments to improve the health of existing hardwood stands and reduce oak mortality.

Callie Schweitzer, research forester with the SRS uplands hardwood unit, coordinates the work among research and national forest personnel setting up the silviculture assessment. This has meant collecting data, selecting stands for the treatments, and keeping the communication going among partners. "Within the next year, the majority of the treatments will be implemented," says Schweitzer. "Then we'll start collecting data on how each treatment affects the structure and species composition of the forest, the ability of oaks to regenerate, and wildlife habitat."

Early into the Fray

Kurt Gottschalk is all for getting the drop on gypsy moth. Research forester with the FS Northern Research Station, Gottschalk has spent over 25 years developing methods to minimize the effects of gypsy moth on oak-dominated forests and to regenerate and rehabilitate them after attacks. He is currently in charge of a Title IV HFRA project on the Monongahela and Wayne National Forests in Ohio, an area where the gypsy moth is already resident or expected within the next couple of years. In an ideal world, Gottschalk would like to see treatments in place 4 to 10 years-or even longer-before infestation.

"Preoutbreak treatments focus on reducing the vulnerability of stands by removing the trees most likely to die and regenerating stands that are close to maturity or understocked," says Gottschalk. "The most effective control is active forest management before the gypsy moth arrives. Everything you do after infestation is reactionary." Gottschalk worked extensively with Schweitzer and others on the Daniel Boone project. "We used some of Kurt's criteria for our initial evaluations of stand health," says Schweitzer. "He also sent his technical staff to the Daniel Boone to help with data collection, and his forester Dave Feicht helped develop marking guidelines for several of the treatments."

The situation at the Daniel Boone is complicated by a high level of oak decline, a naturally occurring process that defoliation by gypsy moths speeds up. One round of defoliation will not usually kill a tree, but trees already weakened by age or drought and then stripped of their leaves can be taken down more easily by other organisms such as shoestring root rot or the twolined chestnut borer.

"You can see gypsy moth as a special case, or subset, of oak decline," says Gottschalk. "In some cases, defoliation by the insect collapses mortality that would occur over 10 years into as little as 2 years." Data from the Northeast and Mid- Atlantic show that about 20 to 25 percent of the areas defoliated will have 50 percent or greater mortality.

Treatment Options

Gottschalk and others have outlined three basic options for dealing with impending gypsy moth infestation: do nothing (the most frequently chosen option); alter the susceptibility of the forest by changing species composition; or alter vulnerability by removing dead and dying trees and improving the vigor of remaining trees by altering stand structure.

The Daniel Boone experiment uses four levels of treatments to both improve vigor and promote oak regeneration. The most intensive treatment, shelterwood harvest with reserves, removes all but the largest trees to eventually create two-aged stands. A second option, oak shelterwood, retains more of the overstory. A third option involves thinning to simply reduce stand density with no special attention given to oak regeneration, while the fourth treatment combines thinning to create a grassy, open woodland habitat with prescribed burning to maintain this condition. Control plots with no treatments, the "do nothing" option, will provide data for comparison.

With over 600 acres of experimental plots mapped out, the silvicultural assessment is one of the largest experiments of its kind and includes multidisciplinary studies on wildlife, stand dynamics using tree-ring data, forest harvesting operations, and the potential of using the wood thinned from plots as biofuel.

"We want to come up with specific recommendations for improving oak regeneration under these situations," says Schweitzer. "We also want to be able to specify the harvesting operations needed to reduce impacts to soil, and the prescribed burning prescriptions that will sustain open oak woodland conditions on these sites."

The experimental plots will also be set up as demonstrations areas, where land managers and private landowners can actually see the effects of different silvicultural approaches on oakdominated forests. Jeff Stringer, associate Extension professor at the University of Kentucky, who has already provided funding for data collection and plot preparation, will provide his expertise in technology transfer when the demonstration plots have been installed.

The enemy is still on the way, but the experiments underway will provide new understanding about how managers can help oak-dominated forests weather the storm. "Compared to other invasive insects, gypsy moth is extremely manageable," says Gottschalk. "If we increase the vigor and regeneration capacity of our forests, they can weather the initial attack with relatively little long-term damage."

Other cooperators include: The National Wild Turkey Federation, Rocky Mountain Elk Foundation, Kentucky Forest Industries Association, Eastern Kentucky University, and the University of Tennessee.

Answers about why that's happening and how to reverse the trend have begun to come in over the past several years. Martin Spetich, SRS upland hardwoods research forester based in Hot Springs, AR, is at work tying together a network of 16 studies centering on upland oak forest dynamics in Arkansas. However, some of the best information comes from two centuries in the past, when Native American populations were prevalent and periodic low-intensity forest fires were a regular occurrence.

There are oaks still standing along the Arkansas River that started growing there between 1719 and 1857, during the time the area was inhabited by the Quapaw Indians. Centuries old, the trees sport scars that show they lived through multiple fires, while their competitors burned away. The lack of fire over the last century of fire suppression may help to explain the decline of oaks in the Ozarks and other areas of Arkansas.

Understanding the role of historic fires in these forests is key to understanding how they developed. To learn more, Spetich teamed up with Richard Guyette at the University of Missouri to document fire history in the Boston Mountains of Arkansas. Together, they have examined fire history back to the early 1600s, and have been able to identify trends in historic fire frequencies associated with such things as human population fluctuations and long-term climate variability.

An Over-Arching View

Part of the problem with studying Arkansas oak forests is that, like most if not all forests, they're constantly evolving, complex systems. The same can be said, though, of Spetich's method of studying them. When he first began working with the Hot Springs unit in 1998, he inherited a number of forest research studies. Spetich calls these studies his orphans.

"They were excellent research studies, but there was really no way to comprehensively tie them together. Without someone to care for them, they could have gone by the wayside like so many other orphaned studies," Spetich said.

Spetich took the orphans under his wing, then started interviewing land managers and others to see what they needed from existing and future studies. The end result was an overarching keystone study that ties together data from 16 integrated projects that span millions of acres and hundreds of years.

"One of the first things I did, back in 1998, was talk to researchers and resource managers around the State, as well as others involved with management of forests, to find out what their real needs were in terms of research information," Spetich says. "There were areas where there was a significant amount of research as well as areas where the data didn't really exist. So I used that information to address their needs by filling in research gaps."

The resulting keystone study examines the effect of fire on species dynamics in the Ozark-Ouachita Highlands, analyzing both burned and unburned areas in combination with a number of treatments. When combined with his other, smaller studies, the keystone study will provide both the bug's-eye view (the opportunity to pinpoint specific stand dynamics and understand interactions with other species) and the bird's-eye view (insight into creating sustainability in Arkansas oaks not only now but across time). It encompasses everything from spatial modeling of large forested landscapes hundreds of years into the future to looking at the ability of a single seedling to survive.

"I wanted to look at the whole Ozark-Ouachita Highlands area across both space and time. There are a number of study sites, and we can work on a portion of them each year. The first part is getting initial measurements; the second part is getting the treatments done. And then there are all of the followup measurements. So each year we get to do something new," Spetich says. "I expect to have preliminary information out in the next 4 years, and then about every 5 years for the next 15 years, when all of this will come together in a comprehensive, integrated way."

From an Acorn

Now, thanks to studies by Loftis, earlier researchers at the Bent Creek Experimental Forest, and many other researchers in the East, we have a much better understanding of the oak regeneration process. The first and most important point is that upland oaks, as a group, will only be present in a new stand if competitive oak regeneration sources were present in the previous stand and persist through the disturbance-in this case, harvest. In its simplest form, the process involves three elements:

• Establishment of new oak seedlings in an existing stand following a good acorn crop
• Development of those seedlings in the existing stand into regeneration sources that can compete successfully when released
• Timely and sufficient release from overstory competition

For oaks to regenerate, all three elements must occur in sequence and within a certain time frame. On many sites in the Southern Appalachians and beyond, the lack of oak regeneration appears to result from the failure of the second element-the development of seedlings into competitive regeneration sources-to occur

"My colleague, retired research forester Paul Johnson with the North Central Research Station (NRS), literally wrote the book on oak ecology and silviculture," says Loftis. "He suggested that on some very dry sites, development of competitive oak regeneration sources can occur without disturbance, because stand conditions are typically sparse enough to provide enough light for seedling development. Paul coined the term accumulator systems to describe this."

But on most other sites, oak seedlings will not develop sufficiently to be able to compete with yellowpoplar and other species after release without some sort of stand disturbance that alters the light regime in just the right way and for a sufficient amount of time prior to release.

In the Southern Appalachians, very shady conditions under mature stands lead to substantial oak seedling mortality and little growth of the surviving oak seedlings. Treatments that provide too much light, while encouraging the growth of oak seedlings, also promote the establishment and development of competitors such as yellow-poplar.

Part of Loftis' research has involved finding a stand treatment that provides the light necessary for oaks to grow, but retards the establishment and development of yellow-poplar. The treatment resulting from this research, sometimes called an oak shelterwood, leaves the main canopy intact but removes much of the vegetation between the main canopy and the ground. After 10 years or so, the small seedlings that had become established as a result of a good acorn crop have developed into much larger seedlings-the regeneration source-now capable of competing after release.

Cross-Region Ambitions

Attendants at a recent SRS All Scientists meeting talked about the possibility of a future cross-regional oak regeneration study. "The objective of the study we are planning is to test oak regeneration prescriptions across upland hardwood forests in the South, and to determine which prescriptions work and where, with a specific focus on the moisture gradient," says Loftis. "Managers need this information to plan appropriate treatments where the regeneration and maintenance of oak ecosystems helps meet their management objectives."

Loftis commented that some suggested prescriptions for regenerating oak include the use of prescribed fire. "Some people suggest that merely ‘returning fire to the ecosystem' will ultimately restore and maintain oak ecosystems, or ecosystems where oak is an important component. Pat Brose (NRS) and Dave Van Lear (Clemson University) developed a much more specific prescription for the use of fire in oak regeneration. I developed a method that does not include the use of fire and seems to be applicable to higher quality sites."

REGENerating More than Oak

Another component of Loftis' research involves the development of regeneration prediction models. His early work focused on predicting the amount of oak that might be expected, if, at a given point in time, an existing stand were regenerated, based on the oak regeneration sources that already existed in that stand.

"The first step in oak regeneration is to assess the adequacy of oak regeneration sources that are present in the existing mature stand where regeneration is contemplated," says Loftis. "Only then will we be able to compare the amount of oak desired in the new stand with what we would expect based on the prediction."

In his more recent work, Loftis has attempted to develop a more general regeneration model.

"I began to think about a general regeneration prediction model in the early to mid-1980s. It is, in some ways, an extension of my work in oak regeneration and the model I developed to assess oak regeneration potential in the Southern Appalachians."

"The earlier model predicts the amount of oak to be expected in a new stand given the oak regeneration sources that are present in the existing mature stand. But on moist sites that support many species, we need a model that predicts species composition-not just the amount of a single species group."

The newer model, a hybrid that includes both an expert systems component and inputs from empirical models, was designed with regeneration harvests in mind, but could be applied to other stand replacing disturbances such as wildfires and major weather events. Known as REGEN, the program that implements the model predicts species composition of a new stand created by substantial overstory removal.

While REGEN was developed for the Southern Appalachian region, it can be adapted for use across most ecosystems and tree species.

"The REGEN program is designed to be modified or calibrated for other areas or ecosystems-wherever the underlying concepts seem to provide a reasonable model of the process, and where scientists and managers feel they have enough information to drive the model," Loftis says. "Currently, a version for western Virginia has been calibrated by Tom Fox at Virginia Tech and for the northern Cumberland Plateau by Jeff Stringer at the University of Kentucky. Within our own unit, Callie Schweitzer has been collecting research data for several years that will provide very important information for calibrating and refining the version for the Cumberland Plateau."

In early June, Loftis announced plans to step down from the project leader position to return his focus to regeneration research, effective this fall.

"I will be attempting to do a synthesis of our older regeneration studies, then use that synthesis to explain the working hypothesis for my regeneration model implemented by the REGEN program. And I will be using some newer data from plots I installed in the 1990s to test and refine REGEN," he wrote in a letter to colleagues.

History has a way of repeating itself. Forest pathologists in the East have reacted with alarm to the threat of a new blight, another nonnative pathogen introduced via nursery stock, this time affecting oaks. The new disease, known as sudden oak death, manifests as bleeding stem cankers, blighted leaves, and twig dieback, and can kill a mature tree in just a few growing seasons. The disease has killed a range of species, including tan oaks and live oaks in California and Oregon. There is no treatment yet available for infected trees; just how the disease unfolds is not completely understood.

PathoProfile

The organism that causes sudden oak death belongs to a class known as the Oomycetes or water molds, a group which includes the potato late blight, the disease responsible for the Irish potato famine in the mid-1800s. Phytophthora ramorum, or Pr for short, gets its genus name from the Greek word for "plant destroyer," and is not to be confused with oak decline, a disease that also affects oak forests in our region. Pr's life cycle includes an airborne phase, where spores float aloft until they land on susceptible hosts, particularly where water is available, perhaps where dew has collected in droplets on a leaf surface or shoot. Pr apparently survives in soil for long periods, probably an adaptation to survive those times when a host isn't present. After penetrating the leaf surface or bark of its host, Pr spreads through host tissue, producing leaf spots, dying twigs, and bleeding cankers on the tree's trunk.

The pathogen was introduced into numerous States in plant nursery shipments, mostly on camellias and rhododendrons, which can serve as sporulators-plants that aren't killed but are sources of reproductive spores. In one of the first known and largest introductions in 2004, 1.5 million potentially infected plants were shipped from a single nursery in California; every State in the United States received potentially infected stock.

Steve Oak, plant pathologist with the Forest Service, Southern Region Forest Health Protection unit and technical coordinator for the sudden oak death early detection survey in U.S. forests, cites his informal scratch-pad estimation that perhaps 2,000 to 5,000 infected plants ended up in people's backyards and gardens nationwide. His nightmare scenario: an infected rhododendron planted in a backyard with a stream nearby, a mature oak forest with a rhododendron understory upwind.

"If eastern oaks are susceptible, we'll have a problem," he says, and so he supplied seedlings of nine compass-august 2007 Steve Oak, forest health plant pathologist, is technical coordinator for the sudden oak death early detection survey in U.S. forests. (Photo by Rod Kindlund, U.S. Forest Service) eastern oak species to be tested with the pathogen at a USDA Agriculture Research Station containment greenhouse at Fort Detrick, MD, by research plant pathologist Paul Tooley. The results showed that all nine eastern oak species tested could be infected under optimal conditions- with the caution that those conditions may be hard to find in nature. The pathogen needs a host to support the production of spores; rhododendron (and perhaps other species) will do, while oaks will not.

Location and Opportunity

For the disease to infect and establish, three conditions must be present: (1) hosts, (2) pathways for introduction, and (3) the right environment. Oak, along with Bill Smith, coordinator with the SRS Eastern Forest Threat Assessment Center, Frank Koch from North Carolina State University, and members of the National Forest Health Monitoring Risk Mapping Team have produced a map of areas where all three factors are present: potential hosts, favorable environment, and pathways-in this case, nurseries that may have received infected plants. The map provides a set of target areas for field monitoring by a team of State forestry agencies trained by Oak in disease recognition and survey protocols. Smith's data show the Southern Appalachians at high risk for infection, along with a small area on the western part of the Mississippi coast.

In addition, Pauline Spaine and Bill Otrosina of the SRS Insects, Diseases, and Invasive Plants unit in Athens, GA, have come up with a descriptive index of some of the microclimate features believed to be critical to an outbreak-a Sudden Oak Death Infection Index-to estimate when and where natural forests could be susceptible to infection. They collected hourly temperature and humidity values in natural forests, choosing sites where previously published data suggested that temperature and humidity values indicated a level of risk for Pr infection. The index can be used to highlight times when the dew point occurs at the optimum temperature for infection for at least 6 hours-times when water would be collecting on leaf surfaces where spores land. Based on their data, Spaine and Otrosina predict that certain areas in western South Carolina and along the Georgia coast, where humidity is high, may be at the greatest risk for Pr infection in the future.

Early Detection is Essential

This spring in Mississippi, rhododendron leaves deployed as Pr bait downstream from a nursery that received infected nursery stock tested positive for the pathogen. The presence of the pathogen is not the same as a disease outbreak. No such outbreak has yet occurred in wild forests in the Eastern United States. While infected vegetation has not been found in the natural environment, State and Federal forest health professionals continue to intensively monitor this site and other areas in the Southeast that might be vulnerable to sudden oak death.

"The State forestry and agriculture agencies are doing a tremendous job with survey and monitoring," says Oak. "Regulation has been improved and strengthened, but it's still leaky. There are two confirmed cases of Pr identified in nurseries in Florida and Mississippi this year so far, but still none found in wild forests anywhere outside of California and Oregon. Still, the takehome message for homeowners and landscapers who want to minimize the risk of introducing Pr is to buy locally propagated plants."

If the disease is found through monitoring, current eradication measures call for ringing the affected area with a 100-foot buffer, then cutting and burning all plant material within the ring. "If it shows up simultaneously in 10 States in 100- acre blocks, there's not much that can be done," says Oak. "There are some key questions still to be answered. What is the latent period between introduction and establishment and between establishment and detectable infestation? How much response time do we have between detection and action?"

Though the disease hasn't been detected in eastern oaks, forest pathologists remain vigilant.

"Are we out of the woods yet? No, the resource at risk is too great to let down our guard," says Oak. "Our only hope is early detection, and it would be irresponsible to assume less than the worst-case scenario-another chestnut blight-until we can get more information about the epidemiology in eastern hosts."

Meanwhile, there's also been a decline in oak regeneration, due to the inability of seedlings that sprout from acorns to grow large enough to be competitive with other species when the overstory on high-quality (moisture, good soil) upland hardwood sites is removed. The problem is that faster growing trees such as maples from stump sprouts and yellowpoplars outgrow the slower growing small oak seedlings. As time goes on, the overstory of the forest becomes dominated by maples or yellowpoplars, and the oaks-which provide Upland Hardwood Forests in Transition food and shelter to a wide range of animals and birds as well as producing high-quality timber-become fewer and fewer.

Oak Regeneration Redux

David Loftis, project leader of the SRS Upland Hardwood Ecology and Management unit (upland hardwoods unit), first came to work at the Bent Creek Experimental Forest in the 1970s. By then, researchers already knew that even though you might have thousands of new oak seedlings from a good acorn crop, less then 5 percent would survive under a dense canopy of overstory trees. They also knew that if you wanted more oaks you had to increase the light reaching the forest floor enough to promote the growth of oak seedlings, but not enough to stimulate sprouting of yellow-poplars from the seed bank. What they didn't know was just how to do this.

Loftis started designing and setting out studies to look at this question in the mid-1970s and early 1980s, and over time has become the acknowledged expert in oak regeneration in the Southern Appalachians. In the past few years, his unit has expanded across the range of upland hardwood forests into the Cumberland Plateau, and even further west into the highlands of Arkansas. The unit has also expanded into wildlife research as managing forests to provide habitat for different creatures has become a major objective for natural resource managers across private, State, and Federal ownerships

In this issue of Compass, we will explore research studies across the range of upland hardwood forests in the South, looking at silvicultural and wildlife research, as well as combinations of the two. We'll look at a project in Kentucky designed to reduce the damage of an insect pest before it ever arrives, and catch up on the latest news and research about sudden oak death. We'll also offer tips about what you as a resource manager or private landowner can do to promote oaks on your own property.

Schweitzer and Stacy Clark, both research foresters with the SRS upland hardwoods unit, are taking us out to three very different sites where they work with a wide range of cooperators to make the scientific connection between silviculture treatments and forest inhabitants-whether they're birds, bats, snakes, frogs, salamanders, or ants. We're starting at a site on the Bankhead National Forest, where a total of 180 plots have undergone 9 different treatments-3 levels of thinning, 3 burning frequencies, and combinations of the two.

The purpose here is to remove the loblolly pines planted in the 1970s and move the forest towards an oak-hickory upland hardwood forest. Schweitzer and Clark are also calibrating the REGEN model- developed by David Loftis to help managers regenerate hardwood forests in the Southern Appalachians-for the Cumberland Plateau. These same plots host studies on a range of different animals and birds carried out by students from nearby Alabama A&M University (AAMU) under funding from SRS and the National Science Foundation, and under supervision of AAMU associate wildlife professor Yong Wang.

A Mantra of Disturbance

Clark is also growing hybrid chestnut seedlings on some of the treatment plots in the Bankhead. One of the things she and Schweitzer want to make clear from the start is what they mean when they talk about restoration and disturbance.

"Trying to reestablish chestnut is an example of true restoration," says Schweitzer. "We are literally trying to restore a tree species that has been extirpated to areas where it once grew. Here, we're trying to restore this forest to a certain structure, but not to what it may have been at some particular point in time, say, before European occupation. We have no way of knowing what that was, but we're pretty sure it wasn't static."

Schweitzer has a mantra of disturbance about the forests of the Cumberland Plateau that's applicable to upland hardwood forests across the South.

"These are disturbance-dependent systems-whether it's the disturbance of early 20th century logging, storms, or insect invasion," she says. "The animals that have taken up residence here are also disturbance dependent. We'll never be able to remove the disturbance from these ecosystems, but we can learn to use silviculture to manipulate disturbance to achieve specific habitat goals. One of the missing pieces is detailed research about animal habitat. That's where the studies by the AAMU students come in."

"But it's silviculture that drives the studies on these sites," she stresses. "We create a disturbance and then look at the effects on specific animals."

One of the plots we visit includes a long aluminum drift fence, really a horizontal funnel that herds snakes and salamanders towards soft traps or white plastic buckets buried in the ground. Graduate student Bill Sutton attaches radio telemetry devices to the snakes he traps here, monitoring their movements during thinning operations and fire. This spring, his bucket traps were filled to the brim with salamanders, hard to believe on this hot, dry day. "There's so much we don't know," says Schweitzer. "From preliminary data, we're seeing that salamanders bury down into the forest floor during prescribed burning, which we've found only affects a thin layer of the duff."

Studies on amphibians and reptiles are particularly important; numbers of these animals have been declining worldwide due to habitat alteration and degradation. Forest management alters habitat, but unlike many other land uses, it allows habitat regrowth. Finding out more about these almost invisible animals can help managers learn to manage for their future.

"The more we know about actual habitat and how different animals respond to management activities, the better we can use management to provide habitat for multiple species," says Schweitzer.

Oaks and Animals

The next day we're on the road again, driving up towards the Tennessee border, to research plots in Jackson County, AL. We're on the top of the Plateau near the Walls of Jericho, a unique formation recently bought by a joint venture between the Nature Conservancy and the Alabama Department of Conservation and Natural Resources. The land the plots are on has been passed back and forth between industry owners, and now belongs to the State of Alabama. We drive down into the forest on a gravel road the locals call "Callie's highway," since it was built to get Schweitzer back into her research sites. On the way down, she points out roadside demonstration sites set up to show different levels of retention cuts. The demonstrations, designed for tours by private landowners and forest managers, are very effective. Looking up from a car, you can easily see the difference between 50 percent retention and a clearcut. That clearcut, by the way, is in its sixth season and already grown up. "The trees were 10 feet tall after the first year," says Schweitzer. "A lot of people wouldn't believe that."

These research sites are Schweitzer's pride and joy, and formed the basis for the studies later installed on the Bankhead National Forest. The road ends, and we walk into a 75-percent retention plot, a pleasant, relatively open forest with a diversity of hardwood species-oak, ash, basswood, sassafras. One of the purposes here is find out how to regenerate oak on good quality sites, where the competition from other tree species such as maple and yellowpoplar is fierce.

The site has some oak seedlings in the understory. The question is how to use forest management to get them to grow up into the overstory, to get the proportion of oak desired for different or multiple purposes. "Here we left the overstory and took out the midstory, except for oak, ash, or persimmon," says Schweitzer. "We followed Loftis' protocol for the midstory treatment, and then installed five different overstory treatments to create five different light levels on the forest floor."

With oak seedlings in the understory, the treatment should have promoted natural oak regeneration- and would have if it weren't for sugar maple. "When they did this experiment at Bent Creek, they didn't have to deal with sugar maple," says Schweitzer.

"Our objective here was to reduce the midstory and allow the small oaks to grow up into a more competitive position. What we did was allow the sugar maples to take the place of the oaks. This brought up the question: What do you do if you don't have the oak seedlings you need for natural regeneration?"

Oaks are advance-regeneration dependent, which means seedlings have to be present on the site before treatment to have a chance of growing into trees. This is in contrast to yellow-poplars and maples, species with a robust seed bank strategy and lightweight seeds easily spread by wind. Schweitzer ended up using artificial regeneration on the site, planting oak seedlings and controlling the midstory competition until, years in the future, the seedlings are competitive enough to be released when the overstory is cut.

There are also chestnut seedlings planted in the understory. Schweitzer and Clark are taking physiological readings on oak and chestnut seedlings under different levels of light on the forest floor, trying to find out where the seedlings use light most efficiently.

Nuthatches and Salamanders

Wang and his students from AAMU have been actively involved on this site studying birds and herpetofauna. They've identified 60 different bird species, and over 40 herpetofauna species, including 11 frogs and toads, 10 salamanders, 5 lizards, 12 snakes, and 2 turtles. One study by graduate student Lisa Barillas involves using radioisotope studies to track the origins of birds that stop over in the area during fall migration. Student Adrian Lesak analyzes songbird community variation among the five levels of overstory retention. For another study, Wang monitored barkforaging and cavity-breeding birds on the five treatment sites, looking at activity in relation to the presence of the dead trees foresters call snags.

"When we took out the midstory, we created a lot of snags out here, which we know are really valuable to wildlife," says Schweitzer. "So we wanted to know if all those snags we created-about 400 an acre, we found out-made any difference to wildlife foraging behaviors. Yong and his students found that barkforaging birds like nuthatches definitely used these sites more than any other. It'll be interesting to follow up with the salamander work."

Managing forests for wildlife depends on learning enough to predict and model habitat, ideally for whole suites of species-and on the larger landscape, for multiple suites of species.

"No species have gone extinct due to forest management," stresses Schweitzer. "With more research, we can learn how to create or maintain good habitat for multiple species. This means doing different things across a landscape, creating a very subtle mosaic of gaps and variations in over-, mid-, and understory that support a wide diversity of plants and animals."

In the scramble to rebuild southern forests after the abusive logging, grazing, and mining of the late 19th century, forest owners and managers lost sight of the lessons learned from their native predecessors. The last century has seen a rebirth of a very different forest than the Europeans inherited. The once dominant chestnut trees have been eliminated by blight. In some places, fire suppression has left large quantities of highly flammable material on forest floors. Dense thickets of rhododendron and mountain laurel block seedling growth of species, like oak, that are preferred by many types of wildlife. In mixed oak-shortleaf pine forests, white pine is replacing the shortleaf component and shading out the oak component. And reduction of available nitrogen- released when nitrogen compounds are exposed to heat-has resulted both in a decline of overstory trees and in fewer and less diverse understory plants.

Solutions Needed

These changes in upland hardwood ecosystems have also opened the way for a host of insects and nonnative invasive plants-and have led to renewed interest in prescribed burning among forestry professionals. Prescribed burning has been used even less in the mountains than in other areas of the South because fire behavior is less predictable and smoke management is more difficult in highly variable topographies. However, the situation has reached a point where the ecological and wildfire-prevention benefits from prescribed burning, combined with advances in burning techniques and smoke management, may cause a shift towards acceptance among upland residents-many more of whom are moving into wildfireprone areas.

Two recent efforts to shift upland forests toward a state that is less stressed and better functioning involve a return to the use of prescribed burning. Both involve research in the mountains of western North Carolina, one at the Coweeta Hydrologic Laboratory near Franklin and the other further east at the Green River Game Land near the South Carolina border. Together they paint a comprehensive picture showing the effectiveness of prescribed burning in reducing wildfire risk and the ecological aftereffects of burning on water quality, availability of nitrogen in soils, competing vegetation, and insects, birds, and small mammals.

Coweeta Study

Since the 1990s, Coweeta scientists have taken a multiple landscape approach to applying prescribed fire of varying sources, intensities, and severity-monitoring the effects on carbon cycling, water quality, and vegetation. Treatments include standreplacement fire (cutting followed by a prescribed burn that simulates wildfire), low-to-moderate intensity understory burning, and wildfire in old-growth hardwoods.

They found that burning successfully reduces invasive white pines and evergreen shrubs such as mountain laurel with little or no movement of sediments into streams, provided the forest floor is kept intact and vegetation recovery is rapid. Based on these findings, the scientists recommend that fire managers should limit fire severity and size if the goal is to minimize the effects of prescribed burning on water quality, soil nutrient loss, vegetation recovery, and other ecosystem properties.

The restoration work at Coweeta is continuing. In ecosystems that have fallen into decline, scientists are evaluating combinations of treatments such as thinning overstory trees and midstory shrubs followed by prescribed fire and the planting of desirable species.

Green River Study

Ten years after the Coweeta studies began, scientists began new research at the Green River Game Land as a part of the National Fire and Fire Surrogate Study, a network of 13 long-term experiments in different settings across the United States. The purpose of the Green River study was to determine the ecological and economic responses of forests to low-intensity prescribed burning with and without thinning and other mechanical treatments that often serve as surrogates for fire.

The Green River study is ongoing, but Tom Waldrop, the SRS research forester who is leading the effort, says that the early results are encouraging. They show that burning and mechanical treatments reduced litter and other forest fuels, which reduce flammability if a wildfire should occur. In addition, organic material in soils increased after prescribed burning. Although mechanical treatments were more effective at removing mountain laurel and rhododendron than fire, oak seedlings from sprouts and acorns increased after prescribed burning.

Jim Hanula, SRS research entomologist based in Athens, GA, reports that an increase in herbaceous plants in the study resulted in increases in a range of insect pollinators including ground nesting bees, wasps and other flies, and even the rare Diana fritillary butterfly, which has been eradicated from parts of its native habitat in North Carolina.

For the same study, Katie Greenberg, research ecologist with the SRS upland hardwoods unit, monitored the effects of treatments on breeding birds, frogs, salamanders, lizards, snakes, white-footed mice, and shrews. She found that low-intensity fires increased populations of whitefooted mice but had no effect on frogs, salamanders, snakes, and shrews. Among the breeding birds, the only species that decreased in numbers were hooded warblers, which make their nests in shrubs, and worm-eating warblers, which forage and make their nests on the ground and in shrubs.

These studies, while not conclusive, suggest that prescribed burning can have a beneficial effect in upland hardwood forests and the creatures that inhabit them.