Richard F. Keim, Jim L. Chambers and Thomas J. Dean

Regional increases in flooding are likely to reduce the productivity of baldcypress-water tupelo swamps in coastal Louisiana. Although these trees are merchantable for lumber production, it will be important to design appropriate management plans for these sites. Given the demand for the wood but reduced site productivity in many natural stands, establishing baldcypress plantations may be a more appropriate long-term strategy for commercial timber production. More baldcypress silviculture research is needed to support intensive management.

More than 800,000 acres of second-growth stands of baldcypress-water tupelo grow in the swamp forests of coastal Louisiana. Most of these stands regenerated naturally after the timber extraction period in the early 20th century and received little subsequent attention. These trees are now large enough to be once again attractive for harvest. People are particularly interested in baldcypress because of the perception that it is naturally rot-resistant. Markets for baldcypress dimensional lumber and garden mulch are developing despite the fact that the second-growth trees are not yet old enough to contain substantial rot-resistant heartwood.

Even as interest in these wood products increases, the swamp forests of coastal Louisiana face difficult times. Largescale changes in hydrological conditions in the region caused by hydrological management and land subsidence have reduced vigor of many stands and impeded regeneration. Therefore, it is increasingly important to understand how swamp forest productivity interacts with flooding, and tools to facilitate management and restoration are needed.

LSU AgCenter scientists in the School of Renewable Natural Resources are addressing these issues with long-term research of how natural stands grow and respond to management in contrasting hydrological conditions. Also, management of many swamp forests in Louisiana for forest products may not be feasible because of reduced productivity and lack of natural regeneration caused by increased flooding. Plantations of baldcypress trees may, therefore, become more important as management of natural stands becomes more focused on wetland values unrelated to timber production, such as wildlife habitat and water quality protection. Few baldcypress plantations have been established, and fewer still have been measured to guide silvicultural prescriptions. The lack of data for baldcypress growth and yield in plantation environments hinders development of effective management plans. To meet these needs, AgCenter researchers are continuing long-term measurements of baldcypress growth in an experimental plantation.

Wetland Forest Productivity and Flooding
To better understand how hydrological changes affect the coastal forest ecosystem, researchers are measuring the width of tree rings in baldcypress at several sites to compare the historical relationship of hydrology and productivity of overstory trees. Sites are in swamps near lakes with long-term records of water levels. One of these research sites (Figure 1) is near Lake Verret. Here flood control levees have eliminated seasonal riverine flooding, but inundation has increased because of subsidence, impeded drainage, and backwater flooding from the outlet of the Atchafalaya Basin.

Growth rates of trees at the Lake Verret site have decreased over the past 30 years, coinciding with increased water levels. Short-term water level changes have different effects on growth than long-term water level changes. For example, trees can grow well when water levels were high for a single year, but the worst growth occurs when water levels are consistently high for several years.

Climate variables such as temperature and precipitation further complicate the picture because they affect growth. These variables often occur in predictable patterns with respect to short-term fluctuations in water levels. Researchers expect to learn enough about the relationships between growth and flooding to predict whether recent declines in baldcypress growth are likely to continue.

Researchers in the School of Renewable Natural Resources also are monitoring long-term research plots in baldcypresswater tupelo forests of the Atchafalaya Basin to understand how differences in flooding affect growth and stand dynamics in a more detailed way than is possible with retrospective analyses of tree rings. In 1980, LSU AgCenter researchers established two sets of plots to compare growth of a stand that experiences flooding early in the growing season to a stand that experiences flooding for nearly all of the growing season. These sites initially had the same volume of tree cover (8,040 cubic feet per acre) (Figure 2). After 25 years, the site with less flooding had grown 27 cubic feet per acre per year to 8,730 cubic feet per acre (Figure 3). The site with more flooding had lost 47 cubic feet per acre per year to 6,863 cubic feet per acre (Figure 4). The volume losses at the more-flooded site occurred because 14 percent of the trees died, and the remaining trees grew very little. In contrast, the less-flooded site experienced 37 percent mortality, but the remaining trees grew enough to more than compensate for the volume loss.

Together, the tree ring and Atchafalaya Basin studies tell a common story: with greater long-term flooding comes less vigorous tree growth. When flooding is not severe, normal forest stand development occurs in which growing space formerly occupied by trees that die is reoccupied through increased growth of the remaining live trees, and forest coverage remains complete. Stands that flood enough to reduce tree growth are on a trajectory to lose contiguous forest cover and eventually convert to treeless marsh or open water.

Baldcypress Management: Experimental Plantation
A study is being conducted on an experimental baldcypress plantation of varying spacing treatments established in the early 1980s. Measurements of the plantation have been made at ages 17 and 22 years (Figures 5 and 6) to better understand how baldcypress responds to competition in a plantation setting. Trees were planted at spacings of 12 feet by 12 feet (289 trees per acre), 16 feet by 16 feet (200 trees per acre), 10 feet by 20 feet (168 trees per acre) and 20 feet by 20 feet (100 trees per acre). There were two 1-acre plots planted at each spacing. The plantation is on a Sharkey clay soil that is poorly drained and marginally productive for many desirable bottomland hardwood species, but suitable for baldcypress.

Results of this project indicate that baldcypress grows just as well at the higher densities as at lower densities for at least the first 22 years. Tree size was about the same regardless of spacing (Figure 7), and mortality averaged 6 percent to 11 percent for all treatments, so high density stands are yielding the greatest timber volumes on a per-area basis (Figure 8).

The main difference among trees grown at different spacings is that trees in denser stands had fewer live branches low on the tree because there was less light. Although this selfpruning is evidence of the beginning of competition among trees in the plantation, that competition has not been sufficient to reduce tree growth. Instead, the higher densities are producing more desirable trees for lumber production because they will have fewer knots in the butt log. Many trees grown in the least dense plots still retain large live branches below 6 feet.

Based on the first 22 years after planting, the management recommendation from this research is to plant baldcypress at densities at least as high as included in this study. Baldcypress is known for its ability to persist at high densities but is also thought to stagnate (decrease in stand-level growth at very high densities). Therefore, more research is needed to learn an appropriate upper limit for baldcypress plantation densities.

Baldcypress Management: Thinning Natural Stands
In conjunction with the Atchafalaya Basin study monitoring the effects of flooding on productivity, LSU AgCenter researchers implemented a study in 1980 to evaluate the response of naturally regenerated baldcypress-water tupelo stands to thinning. At the less-flooded site, stands were thinned to 10 percent to 50 percent of their original density by removing the smaller and poorly-formed trees. Scientists measured these test plots again in 2005 to determine the effects of the thinning on growth of the remaining trees and on the total wood volume produced per area of forest.

As expected, the remaining trees in the thinned plots grew faster than trees in the unthinned plots: 1980-2005 volume growth was 227 cubic feet per acre in the unthinned plots, but 711 cubic feet per acre in the most heavily thinned plots. The largest trees grew in diameter more rapidly after the heavy thin (2.9 inches) than did the trees in the unthinned plots (2.0 inches). However, net stand production from establishment to 2005 was about the same: 2,907 cubic feet per acre in the unthinned plots, and 3,280 cubic feet per acre in the most heavily thinned plots (no statistically significant difference between these values).

This experiment suggests that thinning productive natural stands is an option if the goal includes growing large trees, but thinning is not warranted if simple volume production is desired. The experiment did not include low-vigor stands, and more research is needed to understand how best to manage forests that are subject to prolonged growing-season flooding.

Richard F. Keim, Assistant Professor; Jim L. Chambers, Weaver Brothers Professor of Forestry; and Thomas J. Dean, Professor, School of Renewable Natural Resources, LSU AgCenter, Baton Rouge, La.

(This article was published in the spring 2006 issue of Louisiana Agriculture.)