Hurricane Katrina impact on a leveed bottomland hardwood forest in Louisiana.
Gulf coastal hardwood forests frequently experience disturbance from tropical storms but rarely experience lasting damage due to their long history of coevolution with storms (Conner et al., 1989). Unaltered habitats generally experience only transient storm surge flooding (Conner et al., 1989) and wind damage is highly variable, depending on topography, wind velocity, tree size, and species composition (Everham and Brokaw, 1996). The proportion of individual trees suffering major damage may be very high in sites experiencing the highest wind velocities, but outright mortality is less common (Table 1); and most damaged trees regrow rather than die (Everham and Brokaw, 1996). Regrowth of damaged trees and release of understory seedlings and saplings after disturbance (Boucher et al., 1990; Tanner et al., 1991; Yih et al., 1991; Basnet, 1993; Boucher et al., 1994; Lugo and Scatena, 1996) often promotes long-term stability of hardwood forest community composition despite hurricane damage (Batista and Platt, 2003; Keeland and Gorham, 2009).
Previous studies of hurricane impacts have focused on forests relatively unaffected by human activities, because study sites have usually been established with the intention of understanding natural processes that regulate community composition and ecosystem properties. However, bottomland hardwood forests are increasingly altered by development such as levee construction, drainage, and river channelization (Gore and Shields, 1995; Fredrickson, 1997). These changes may fundamentally alter the interaction of hurricanes with coastal forests, potentially increasing their susceptibility to major change in species composition and invasion by non-native species. These effects are well-documented for human-altered coastal marshes (Conner et al., 1989), but less is known about bottomland hardwoods. The possibility of increased future hurricane frequency and intensity makes understanding the interaction of floodplain development and hurricane disturbance critical to the management of these habitats.
The purpose of this study was to evaluate the impact of Hurricane Katrina on a tract of bottomland hardwood forest on the Bayou Sauvage National Wildlife Refuge (BSNWR). This forest tract is cut off from natural sources of accretion by levees, which trapped the storm surge from Hurricane Katrina and exposed the forest to saline water much longer than is usual in hurricanes. The site thus offers the opportunity to compare hurricane impacts on an area with extensive human alteration of natural hydrological patterns to impacts in similar but substantially unaltered bottomland forests (Table 1). I studied woody vegetation on permanent plots in the BSNWR forest to determine (1) forest composition and structure immediately before the storm, (2) the overall impact of the storm on the woody plant community, and (3) how the current and projected distributions of the invasive tree Triadica sebifera were affected by the interaction between hurricane disturbance and human alteration of natural hydrological patterns.
Study site.--The study site occupies the natural levee of Bayou Sauvage, an ancient distributary of the Mississippi River that formerly ran 50 km from present-day Kenner, LA east to the Chef Menteur Pass (Spearing, 1995). The area has been modified by sugarcane cultivation in the 19th century (Penfound and Howard, 1940), logging of Taxodium distichum (L.) L.C. Rich. from adjacent swamps (Wall and Darwin, 1999), and most recently by extensive levee construction to enhance the quality of surrounding marsh habitat for migratory waterfowl (J. Harris, pers. comm.). This forest has been surveyed several times since the abandonment of agriculture in 1886, and these studies have documented the regrowth and maturation of secondary forest and the invasion of the introduced Chinese tallow tree, Triadica sebifera (L.) Small (Penfound and Howard, 1940; Wall and Darwin, 1999; White and Skojac, 2002).
The BSNWR forest occupies approximately 100 ha in a strip 300-500 m wide near the junction of U.S. Highways 90 and 11, between Bayou Sauvage on the south and Blind Lagoon on the north (Fig. 1). Prior to the storm the area supported a well-developed bottomland hardwood forest approximately 120 y old, growing on Sharkey clay (Trahan, 1989). The study site was located on the northern end of the tract, corresponding to transects 3-6 in Wall and Darwin (1999). The area is characterized by low vertical relief, with an elevation change of 0.5-0.7 m between the natural levee of Bayou Sauvage and the edge of Blind Lagoon. The entire area is cut off from sources of accretion by levee construction, and the lowest portions of the ridge forest adjacent to Blind Lagoon have subsided below sea level. Rain is the source of nearly all of the water on the refuge, and water in Blind Lagoon was essentially fresh prior to the storm (U.S. Fish and Wildlife Service, 2009).
Hurricane Katrina.--Katrina made landfall on 29 Aug. 2005 at the mouth of the Pearl River, on the boundary between Louisana and Mississippi 30 km east of Bayou Sauvage (Fig. 1). The outer eyewall passed directly over BSNWR, which experienced sustained winds of 120 kph and gusts exceeding 190 kph (Knabb et al., 2006b). In addition, easterly winds pushed a 5-6 m storm surge over the levee and into the refuge from the adjacent Mississippi River Gulf Outlet, a waterway that connects the inner harbor complex in eastern New Orleans directly to the Gulf of Mexico (Fig. 1). This storm surge completely covered the refuge and remained trapped there for 4 wk until the U.S. Fish and Wildlife Service was able to open weirs to drain the water (U.S. Fish and Wildlife Service, 2009). The salinity of the Lake Borgne, which is immediately to the east of the refuge, is typically 10-15 ppt (Orlando, 1993); however, evaporation subsequently increased salinity to 17-20 ppt in floodwaters standing on the refuge at the time of drainage (U.S. Fish and Wildlife Service, 2009). Blind Lagoon retained some elevated residual salinity in the water column after drainage, estimated at 5 ppt in 2009 (U.S. Fish and Wildlife Service, 2009).
Vegetation surveys.--Twelve plots, each 0.25 ha, were established along the elevational gradient from the top of the back ridge along Bayou Sauvage, avoiding the sharp elevation change at the bayou edge, to the limit of the forest near the edge of Blind Lagoon. Woody plants and palmettos (Sabal minor (Jacq.) Pers.) were surveyed between May and Sept. 2004 in circular quadrats 10 m in radius at the center of each plot. Center points were marked with a length of PVC pipe driven into the soil. GPS was used to establish coordinates of plot centers. The elevation of plot center points was estimated relative to one another and to a nearby USGS vertical control point (Orleans Parish J276, PID: BHll08) using a digital transit. Due to the subsidence of the general landscape in south Louisiana, relative elevations of plot centers were used in this study only to estimate depth of inundation on each plot (see below).
All palmettos, and all trees and shrubs 2 m or more in height, were identified to species using Radford et al. (1968). DBH was measured for trees and shrubs during Oct. 2004 but was omitted for Sabal minor due to the difficulty of obtaining accurate measurements on horizontal stems. Although Wall and Darwin (1999) recorded Ulmus rubra Muhl. as the only member of this genus on the site, I found some individuals of Ulmus to be ambiguous in morphology, more closely resembling U. americana L. in sterile condition. This ambiguity was not resolved prior to the hurricane, and this species was simply classified as Ulmus spp. for this analysis. The position of each plant relative to the center point (distance, bearing) was recorded with the intention of giving each a permanent tag during 2005. However, only a few large trees had been tagged prior to the storm in Aug. 2005.
[FIGURE 1 OMITTED]
Data collected were used to estimate density (individuals x [ha.sup.-1]) and dominance according to basal area ([m.sup.2]x[ha.sup.-1]) for each species on each plot. These measures were used to calculate relative density (stems/total stems), relative dominance (basal area/total basal area), and relative frequency (frequency of plots on which individuals were found) (Barbour et al., 1987). Importance value was calculated as the sum of relative density, relative dominance, and relative frequency (Barbour et al., 1987). Measures requiring DBH (basal area, relative dominance, importance value) were not calculated for palmetto.
Initial visits to the study site after Hurricane Katrina occurred after leaf flush, during Apr.-Jun. 2006, and focused on re-establishing plot centers, as all center markers had been moved, buried, or lost entirely. GPS was first used to establish the approximate location of plot centers, which were then resurveyed exactly by triangulation from the large or distinctive trees on each plot. Resurvey of the plots was carried out from Jul.-Sept. 2006, following the methods outlined above. Each tree was classified as living, uprooted, snapped off, or standing dead. No quantitative assessment of defoliation or limb loss was made due to the elapsed time between the hurricane and the re-survey. Density, dominance, relative density, relative dominance, relative frequency, and importance value were calculated for all remaining living plants using the methods outlined above.
[FIGURE 2 OMITTED]
Although no estimates of flood depth were made immediately after the storm, floodwaters left clearly visible evaporation rings on standing stems (Fig. 2). These rings were not present prior to Hurricane Katrina (Fig. 2), as water on the refuge is regulated at levels well below the elevation of the ridge forest (U.S. Fish and Wildlife Service, 2009). These rings provided a way to retrospectively estimate the depth of standing water on each plot between the passage of the storm on 29 Aug. and the subsequent drawdown of water. Several rings were visible, with the highest deposit likely corresponding to the level of transient flooding that occurred in the immediate wake of the storm. I used the top of the heaviest evaporation ring as a benchmark to estimate the maximum standing water depth during the inundation period. Evaporation rings were not clearly visible on the highest elevation plots, indicating that flood levels were at or near the soil surface, or that no standing water had been present. For these plots I used relative elevation change from nearby plots on which water depth was measured from distinct evaporation rings to estimate the depth of flooding after the storm.
During Oct. 2006, 49 seedling plots, each 2 X 2 m, were surveyed and all seedlings of woody plants identified to species. One set of 25 plots was established at random positions along a transect through the center of the study site from Bayou Sauvage to Blind Lagoon, and two seedling plots were surveyed at random points on each of the 0.25 ha plots. Seedling plot elevations were estimated by reference to the nearest plot center using transit and rod.
Statistical analysis.--All plot-level species counts were non-normally distributed, and no transformation succeeded in producing data that met assumptions of standard parametric tests. For species found on five or more plots, I used the binomial test to evaluate the significance of change in species counts on plots before and after the storm (Zar, 1999). I compared the number of plots on which species counts decreased to those on which species counts either increased or were unchanged.
Prior to Hurricane Katrina, 1923 individual plants of 11 species were found in the total surveyed area of 3770 [m.sup.2]. Of these, 187 individuals in eight species were classified as overstory trees with a diameter of 10 cm or more, with a density of 494.5 [+ or -] 180.0 stems-[ha.sup.-1] (Table 2). Shannon diversity of the overstory was calculated at 0.729. The most common overstory tree was the introduced species Triadica sebifera (n = 57) and the most common native was Celtis laevigata Willd. (n = 47); together, these two species accounted for nearly 70% of all stems in the overstory.
Density of understory plants (1-10 cm DBH) was 4595 [+ or -] 1347 stems x [ha.sup.-1]; of these, most were palmettos, which occurred with a density of 2835 [+ or -] 1347 stems x [ha.sup.-1]. Shannon diversity in the understory was lower than in the canopy (0.499 vs. 0.729), reflecting the dominance of Sabal minor. Eight woody dicot species also occurred in the understory, with a combined density of 1760 stems-[ha.sup.-1] (Table 2). Among these eight species Triadica sebifera was numerically more dominant in the understory than the overstory, accounting for over 60% of all woody stems, and Shannon diversity of understory trees and shrubs (excluding S. minor) was only 0.538. Ilex decidua Walt. was the most common native understory shrub, and Acer rubrum L. and Ulmus spp. were the most common native trees. Although Quercus nigra L. and Q. virginiana P. Mill. occurred in the overstory, no individuals were found in the understory.
Excluding Sabal minor, Triadica sebifera was numerically dominant in both overstory and understory and had the highest relative density of any species (Table 2). However, Celtis had almost twice the basal area of Triadica, and the combined basal area of two oaks, Quercus nigra and Q. virginiana, also exceeded that of Triadica. Celtis and Q. virginiana had the highest relative dominance among native trees due to the large diameter of the surveyed individuals, while Triadica also had high relative dominance due to the large number of small individuals surveyed. Triadica had the highest importance value overall, due to the large number of small individuals found on all plots, while Celtis had the highest importance value of native species due to its large mean diameter and wide occurrence on the plots. Sabal minor was the most abundant plant overall, occurring on all 12 plots.
Of the plants surveyed, 612 remained alive 1 y after the storm in Sept. 2006, for a total mortality rate of 68.2% (Table 3). Two species, Diospyros virginiana and Ilex vomitoria Ait., were entirely eliminated from the 12 survey plots. Little damage from wind was evident; only 15 trees of three species were clearly uprooted during the storm, accounting for slightly over 1% of total stems. Except for these 15 individuals, all other dead trees on the plots were still standing in Sept. 2006. Trees uprooted by the storm did not differ in DBH from the prestorm DBH distribution in any of the three species (Celtis: t = 1.625, df = 61, P = 0.11; Ulmus: (t = 1.77, df = 67, P = 0.081; Triadica: t = 0.24, df = 481, P = 0.81). None of the study trees were snapped off.
Nearly all mortality on the study site was due to inundation with floodwater (Table 3). Evaporation rings were visible on virtually all standing dead trees, whereas their major branches were largely intact. Overall mortality of plants showed a significant, non-linear relationship to standing water depth estimated from evaporation rings (P < 0.025; [r.sup.2] = 0.41; Fig. 3).
Storm mortality dramatically altered the composition of the plant community (Table 4). Only 43 individuals of four overstory species survived the storm, and overstory density was reduced by over 75% to 114.0 [+ or -] 105.9 stems x [ha.sup.-1]. Triadica sebifera increased from 30% of overstory stems before the storm to 44% after, and Quercus virginiana replaced Celtis laevigata as the most common native tree in the overstory. The increased importance of T. sebifera and loss of native species due to the storm reduced overstory diversity from 0.729 to 0.588. Binomial tests revealed significant decreases (P < 0.05) in abundance of overstory C. laevigata, Liquidambar styraciflua, and T. sebifera but not Q. virginiana (P = 0.50).
[FIGURE 3 OMITTED]
Only the five most common species prior to the storm survived in the understory; overall density was reduced by nearly 75% to 1475 [+ or -] 1708 stems x [ha.sup.-1] and Shannon diversity was reduced to 0.345 (excluding Sabal minor. 0.245) (Table 4). Binomial tests showed that five species, Acer rubrum, Ilex decidua, S. minor, Triadica sebifera and Ulmus spp., suffered significant (P < 0.05) reductions in density due to the storm. Sabal minor and T. sebifera remained the most numerous species in the understory while A. rubrum, I. decidua and Ubmus spp. persisted but at greatly reduced densities (Table 4). Three species, Diospyros, I. vomitoria, and Celtis laevigata, were entirely eliminated from the understory.
Basal area in the ridge forest was reduced by over 70% after the storm, from 26.5 to 7.7 [m.sup.2] [ha.sup.-1] (Table 4). Only Quercus virginiana, which was found on the higher plots, maintained more than half its pre-storm basal area. Triadica lost almost 75% of basal area and several other species lost at least 90% of their pre-storm basal area. The reduction in Celtis basal area was particularly dramatic, accounting for over 40% of the total reduction in basal area on the study site. Although surviving individuals of Celtis were large, its importance value dropped to second among native trees. Triadica and Q. virginiana increased in relative dominance, largely due to the loss of basal area of Celtis and to a lesser extent Q. nigra and Ulmus (Table 4). Triadica retained its high importance value in the community after the storm, but nearly all other species showed sharp declines in importance value. The only native species to increase in importance value after the storm was Quercus virginiana, which was characterized by large individuals that experienced relatively low mortality. After the storm Q. virginiana replaced Celtis as the native species with the highest importance value (Table 4).
Seedlings of Triadica sebifera, Celtis laevigata, Ilex decidua, and Acer rubrum were found on the 49 seedling plots. Triadica was the most abundant species, with a density seven times that of Celtis, the next most abundant species, and four times that of all native seedlings combined (Fig. 4).
The observed mortality rate of 68.2% in the ridge forest of Bayou Sauvage is the highest ever documented from a Gulf Coast hardwood forest hit by a hurricane. Tree mortality from Hurricane Katrina in the Pearl River basin was substantially lower (Chapman et al., 2008), suggesting that site-specific factors, particularly impoundment of floodwaters by levees, were responsible for high levels of mortality at Bayou Sauvage. Post-storm surveys clearly show that the high level of damage to the existing plant community has opened the way for expansion by non-native species such as Triadica sebifera.
[FIGURE 4 OMITTED]
Pre-Katrina vegetation patterns.--The vegetation of the study site prior to Hurricane Katrina was consistent with previous recent surveys of BSNWR and other forest remnants south of New Orleans (Wall and Darwin, 1999; White and Skojac, 2002) and the few discrepancies were mostly attributable to minor differences in methods. This study sampled a smaller area and a smaller range of elevations than Wall and Darwin (1999) and so lacked many of the minor species found by them on the highest elevations in this forest. Importance values in this study differed substantially from those reported in Wall and Darwin (1999), perhaps because they used smaller survey plots (10 [m.sup.2]) in which large trees were less likely to occur. Although clearly threatened by encroaching Chinese tallow tree, prior to the storm the ridge forest still retained a substantial native component with a canopy dominated by Celtis laevigata and an understory dominated by Sabal minor.
Hurricane impacts--Although some non-storm related mortality would be expected between surveys, annual mortality in similar forests is generally 1-2% (Harcombe et al., 2002; Batista and Platt, 2003; Chapman et al., 2008; Keeland and Gorham, 2009), suggesting that nearly all trees in this study died as a result of Hurricane Katrina. Although damage from the eyewall winds of Hurricane Camille (1969) reached a similar level (Gunter and Eleuterius, 1973), many trees blown down by Camille likely resprouted rather than died. The mortality observed in this study is thus unusual both for its magnitude and because it was not associated with extensive blowdown damage as is typical in windstorms (Everham and Brokaw, 1966).
Bayou Sauvage experienced sustained winds in excess of 120 km x [hr.sup.-1] with higher gusts, but less than 1% of all plants died as a result of wind damage. This low level of mortality is consistent with the relatively low mortalities measured in other bottomland sites experiencing winds of similar intensity (Table 1). Unlike previous studies that found taller trees at higher risk of wind throw (Gresham et al., 1991; Putz and Sharitz, 1991; Walker, 1991; You and Petty, 1991), there was no consistent relationship between height and toppling. Most uprooted trees were relatively small understory Triadica that did not exceed 8 m in height, and most were on a single plot in close proximity, suggesting that they were blown down by a particularly strong gust. The lack of relationship between tree height and wind damage is not completely unexpected; Herbert et al. (1999) found no relationship between uprooting and tree height in a Hawaiian Metrosideros forest after Hurricane Iniki.
Nearly all mortality from the hurricane appears to have been caused by inundation by floodwater driven onto the refuge by storm surge from the adjacent Mississippi River Gulf Outlet. The observed mortality was over three times that found in unleveed forests of the Pearl River Basin, which experienced much higher sustained winds during Hurricane Katrina (Chapman et al., 2008). Although it is not entirely clear how long each plot remained inundated, the relationship of mortality and water depth suggests that duration of inundation was likely the crucial factor in mortality. Mortality was exacerbated by the construction of levees that impounded floodwaters on the refuge for up to 4 wk, and in this sense human alteration of the natural hydrology of this area was directly responsible for most of the damage to the vegetation of the refuge forest by the hurricane. Had storm surge followed the natural hydrological pattern of inundation followed by rapid drainage after the departure of the hurricane it seems likely that the BSNWR forest would have experienced little or no disturbance from Hurricane Katrina.
Although bottomland hardwood tree species are known to vary widely in their sensitivity to flooding and salinity (Pezeshki et al., 1990; Connor and Askew, 1993) the duration of flooding from Hurricane Katrina's storm surge was long enough to severely affect even relatively tolerant species. Acer rubrum, Liquidambar styraciflua, Celtis laevigata and Quercus nigra are known to be intolerant of short-term freshwater or saline storm surge flooding (Conner and Askew, 1993; Conner and Inabinette, 2003; Williams et al., 1998) and high mortality rates observed in these species are not surprising. In contrast, Triadica sebifera seedlings are capable of surviving up to 6 wk of inundation in 10 ppt saline water (Conner, 1994), and high mortality in this species is likely due to excess salinity of floodwaters. Even Quercus virginiana, which is known to tolerate up to 6 mo of exposure to salinities up to 22 ppt (Williams et al., 1998), suffered a 33% mortality rate, even though this was substantially lower than other tree species on the five plots on which it occurred (C. laevigata: 90%; L. styraciflua: 100%; Ulmus spp.: 100%; T. sebifera; 89%).
Hurricane Katrina storm surge flooding also affected a wider range of size classes than expected from wind damage alone. Flood mortality in this study was essentially identical among overstory and understory plants, in marked contrast to the typical pattern of hurricane wind mortality increasing with tree diameter (Foster, 1988; Gresham et al., 1991; Putz and Sharitz, 1991; Walker, 1991; You and Petty, 1991; Batista et al., 1998). Pervasive flood mortality thus removed a large fraction of the BSNWR plant community and likely shifted the regeneration pathway from regrowth of damaged trees and advanced regeneration toward new recruitment from the seed bank.
Distribution and Abundance of Triadica setnfera.--The increase in dominance of Triadica sebifera in the post-Katrina forest appears to be an unintended consequence of levee construction. Though greatly reduced in density after the storm, overstory Triadica comprised the largest pool of reproductively mature trees on the entire study site, outnumbering native species on every plot. Understory Triadica was also more dominant than before the storm and had a more favorable competitive environment due to the loss of the native canopy.
Evidence suggests that storm surge flooding also facilitated recruitment of new Triadica sebifera at BSNWR. Seedling plot surveys suggest that the seed bank was dominated by Triadica, and the increased light and nutrient availability expected from hurricane damage (Brokaw and Greer, 1991; Fernandez and Fetcher, 1991; Silver et al., 1996; Lugo and Frangi, 2003) coupled with reduced competition from native species would be expected to provide favorable conditions for Triadica recruitment (]ones and McLeod, 1990; Jones and Sharitz, 1990; Jones, 1993). In contrast, poor recruitment of new seedlings was observed for the native species that dominated the canopy prior to the storm. The lack of recruitment of native trees into stands of Chinese tallow in Texas coastal prairie has been attributed to seed limitation (Siemann and Rogers, 2006), but this is not the case at Bayou Sauvage. Annual seed inputs on the study site prior to the storm averaged at least 300 [m.sup.2] for both Celtis laevigata and Ulmus spp., and 50 m 2 for Acer rubrum (Howard, unpubl.), and yet recruitment of these species after the storm was far lower than that of Triadica. A paucity of seedlings of canopy dominants is common in bottomland forests (Harcombe and Marks, 1977; Levy and Walker, 1979; Streng et al., 1989; Jones et al., 1994; White and Skojac, 2002) and may be related to the episodic occurrence of conditions favorable for establishment of floodplain tree species (Levy and Walker, 1979; Conner et al., 1986; Streng et al., 1989). In contrast, the persistent seed bank of Triadica (Cameron et al., 2000) permits it to respond quickly to major disturbances, and the high growth rate of this species suggests that it poses a risk to become even more dominant in the Bayou Sauvage ridge forest. In response to the prospect of increased dominance of Triadica in the wake of Hurricane Katrina, the refuge management plan has proposed increased frequency of active control measures (U.S. Fish and Wildlife Service, 2009).
Hurricane disturbance has been shown to promote invasion of native communities by exotic invasive trees (Conner et al., 2002; Bellingham et al., 2005), and extensive windthrow in the Pearl River basin during Hurricane Katrina has facilitated increased recruitment of Triadica sebifera (Chapman et al., 2008). However, the impact of Katrina on BSNWR suggests that floodplain development in the Gulf and South Atlantic coastal plains may amplify hurricane disturbance even in areas not directly affected by catastrophic wind damage. Levees in coastal habitats across the southeastern United States may have made much of this area potentially vulnerable to the same kind of community turnover now expected at BSNWR. Floodplain development may interact with eustatic sea level rise and the possibility of increased frequency and intensity of tropical storm systems to produce abrupt and unexpected shifts in species composition, frequency of invasive species, and ecosystem function in human-altered plant communities throughout the southeastern United States.
Acknowledgments.--I thank the U.S. Fish and Wildlife Service Southeastern Louisiana Refuges, particularly J. Harris, for permission to work at Bayou Sauvage and for their generous support for all phases of this project. K. Law, M. Wroten and L. Hibdon assisted with plot establishment and pre-storm surveys. E. Estevez, A. Skinner, S. Brown, and the 2006 cohort in the Undergraduate Mentoring in Environmental Biology program at UNO assisted with post-storm surveys. I am grateful for comments from two anonymous reviewers, which improved the manuscript. I am grateful for financial support from the Coypu Foundation and the College of Sciences at the University of New Orleans.
BARBOUR, M. G., J. H. BURK, AND W. D. PITTS. 1987. Terrestrial Plant Ecology, 2nd ed. Benjamin Cummings, Menlo Park, California. 634 p.
BASNET, K. 1993. Recovery of a tropical rain forest after hurricane damage. Vegetatio, 109:1-4.
BATISTA, W. B. AND W. J. PLATT. 2003. Tree population responses to hurricane disturbance: syndromes in a south-eastern USA old-growth forest. J. Ecol., 91:197-212.
--, --, AND R. E. MACCHIAVELLI. 1998. Demography of a shade tolerant tree (Fagus grandifolia) in a hurricane disturbed forest. Ecology, 79:38-53.
BELLINGHAM, P. J., E. V. J. TANNER, AND J. R. HEALEY. 2005. Hurricane disturbance accelerates invasion by the alien tree Pittosporum undulatum in Jamaican montane rain forests. J. Veg. Sci., 16:675-684.
BOUCHER, D. H., J. H. VANDERMEER, M. A. MALLONA, N. ZAMORA, AND I. PERFECTO. 1994. Resistance and resilience in a directly regenerating rainforest: Nicaraguan trees of the Vochysiaceae after Hurricane Joan. For. Ecol. Manage., 68:127-136.
--, --, K. YIH, AND N. ZAMORA. 1990. Contrasting hurricane damage in tropical rain forest and pine forest. Ecology, 71:2022-2024.
BROKAW, N. V. L. AND J. S. GREAR. 1991. Forest structure before and after Hurricane Hugo at three elevations in the Luquillo Mountains, Puerto Rico. Biotropica, 23:386-392.
CAMERON, G., E. G. GLUMAC, AND B. D. ESHELMAN. 2000. Germination and dormancy in seeds of Sapium sebiferum (Chinese tallow tree). J. Coastal Res., 16:391-395.
CHAPMAN, E. L., J. Q. CHAMBERS, K. F. RIBBECK, D. B. BAKER, M. A. TOBLER, H. ZENG, AND D. A. WHITE. 2008. Hurricane Katrina impacts on forest trees of Louisiana's Pearl River basin. For. Ecol. Manage., 256:883-889.
CONNER, W. H. 1994. The effect of salinity and waterlogging on growth and survival of baldcypress and Chinese tallow seedlings. J. Coast. Res., 10:1045-1049.
-- AND G. R. ASKEW. 1993. Impact of saltwater flooding on red maple, redbay, and Chinese tallow seedlings. Castanea, 58:214-219.
--, J. W. DAY, R. H. BAUMANN, ANDJ. M. RANDALL. 1989. Influence of hurricanes on coastal ecosystems along the northern Gulf of Mexico. Wetl. Ecol. Manage., 1:45-56.
-- AND L. W. INABINETTE. 2003. Tree growth in three South Carolina (USA) swamps, after Hurricane Hugo: 1991-2001. For. Ecol. Manage., 182:371-380.
--, I. MIHAL1A, AND J. WOLFE. 2002. Tree community structure and changes from 1987 to 1999 in three Louisiana and three South Carolina forested wetlands. Wetlands, 22:58-70.
--, J. R. TOLWER, AND F. H. SKLAR. 1986. Natural regeneration of baldcypress (Taxodium distichum (L.) Rich.) In a Louisiana swamp. For. Ecol. Manage., 14:305-317.
EWRHAM, E. M. AND N. V. L. BROKAW. 1996. Forest damage and recovery from catastrophic wind. Bot. Rev., 62:113-185.
FERNANDEZ, O. S. AND N. FETCHER. 1991. Changes in light availability following Hurricane Hugo in a subtropical montane forest in Puerto Rico. Biotropica, 23:393-399.
FOSTER, D. 1988. Species and stand responses to catastrophic wind in central New England, U.S.A. J. Ecol., 76:135-151.
FREDRICKSON, L. H. 1997. Managing forested wetlands, p. 147-177. In: M. R. Boyce and A. Haney (eds.). Ecosystem management: applications for sustainable forest and wildlife resources. Yale University Press, New Haven, Connecticut. 384 p.
GORE, J. A. AND F. D. SHIELDS, JR. 1995. Can large rivers be restored? Bioscience, 45:142-152.
GRESHAM, C. A., T. M. WILLIAMS, AND D. J. LIPSCOMB. 1991. Hurricane Hugo wind damage to southeastern U.S. coastal tree species. Biotropica, 23:420-426.
GUNTER, G. AND L. N. ELEUTERIUS. 1973. Some effects of hurricanes on the terrestrial biota, with special reference to Camille. Gulf Res. Reports, 4:174-185.
HARCOMBE, P. A., C. J. BILL, M. FULTON, J. S. GLITZENSTEIN, P. L. MARKS, AND I. S. ELSIK. 2002. Stand dynamics over 18 years in a southern mixed hardwood forest, Texas, U.S.A. J. Ecol., 90:947-957.
--, L. E. MANN LEIPZIG, AND I. S. ELSIK. 2009. Effects of Hurricane Katrina on three long-term forest study plots in east Texas, USA. Wetlands, 29:88-100.
-- AND P. L. MAP, IS. 1977. Understory structure of a mesic forest in southeast Texas. Ecology, 58:1144-1151.
HERBERT, D. A., J. A. FOWNES, AND P. M. VITOUSEK. 1999. Hurricane damage to a Hawaiian forest: nutrient supply rate affects resistance and resilience. Ecology, 80:908-920.
JONES, R. H. 1993. Influence of soil temperature on root competition in seedlings of Acer rubrum, Liquidamabar styraciflua, and Sapium sebiferum. Am. Midl. Nat., 130:116-126.
--AND K. W. MCLEOD. 1993. Growth and photosynthetic responses to a range of light environments in Chinese tallowtree and Carolina ash seedlings. For. Sci., 36:851-862.
-- AND R. R. SHARITZ. 1990. Effects of root competition and flooding on growth of Chinese tallow tree seedlings. Can. J. For. Res., 20:573-578.
--, --, P. M. DIXON, D. S. SEGAL, AND R. L. SCHNEIDER. 1994. Woody plant regeneration in four floodplain forests. Ecol. Monogr., 64:345-367.
KEELAND, B. D. AND L. E. GORHAM. 2009. Delayed tree mortality in the Atchafalaya Basin of southern Louisiana following Hurricane Andrew. Wetlands, 29:101-111.
KNABB, R. D., D. P. BROWN, AND J. R. RHOME. 2006a. Tropical Cyclone Report: Hurricane Rita. http://www.nhc.noaa.gov/pdf/TCR-AL182005_Rita.pdf
--, J. R. RHOME, AND D. P. BROWN. 2006b. Tropical Cyclone Report: Hurricane Katrina. http://www.nhc.noaa.gov/pdf/TCR-AL122005_Katrina.pdf
LEVY, G. F. AND S. W. WALRER. 1979. Forest dynamics in the Dismal Swamp of Virginia, p. 101-126. In: P. W. Kirk, Jr. (ed.). The Great Dismal Swamp. University of Virginia Press, Charlottesville, Virginia. 439 p.
LUGO, A. E. AND J. L. FRANGI. 2003. Changes in necromass and nutrients on the forest floor of a palm floodplain forest in the Luquillo Mountains of Puerto Rico. Carib. J. Sci., 39:265-272. -- AND F. N. SCATENA. 1996. Background and catastrophic tree mortality in tropical moist, wet, and rain forests. Biotropica, 28:585-599.
NOAA. 2002a. Tropical Cyclone Report: Kate. http://www.nhc.noaa.gov/archive/storm_wallets/ atlantic/at11985-prelim/kate/.
--. 2002b. Tropical cyclone report: Bonnie. http://www.nhc.noaa.gov/archive/storm_wallets/ atlantic/aft1986-prelim/bonnie/.
ORLANDO, S. P. 1993. Salinity characteristics of Gulf of Mexico estuaries. NOAA Strategic Environmental Assessments Division, Office of Ocean Resources Conservation and Assessment, Washington, D.C. 197 p.
PENFOUND, W. T. AND J. R. HOWARD. 1940. A phytosociological study of an evergreen oak forest in the vicinity of New Orleans, Louisiana. Am. Midl. Nat., 23:165-174.
PEZESHKI, S. R., R. D. DELAUNE, AND W. H. PATRICK. 1990. Flooding and saltwater intrusion - potential effects on survival and productivity of wetland forests along the United States guff coast. For. Ecol. Manage., 33/34:287-301.
PUTZ, F. E. AND R. R. SHARITZ. 1991. Hurricane damage to old-growth forest in Congaree Swamp National Monument, South Carolina, U.S.A. Can. J. For. Res., 21:1765-1770.
RADFORD, A. E., H. E. AHLES, AND C. R. BELL. 1968. Manual of the vascular flora of the Carolinas. University of North Carolina Press, Chapel Hill, North Carolina. 1183 p.
RAPPAPORT, E. 2005. Tropical Cyclone Report: Hurricane Andrew. http://www.nhc.noaa.gov/1992andrew.html.
SIEMANN, E. AND W. E. ROGERS. 2006. Recruitment limitation, seedling performance, and the persistence of exotic tree monocultures. Biol. Invas., 8:979-991.
SILVER, W. L., F. N. SCATENA, A. H. JOHNSON, T. G. SICCAMA, AND F. WATT. 1996. At what temporal scales does disturbance affect belowground nutrient pools? Biotropica, 28:441-457.
SPZARING, D. 1995. Roadside Geology of Louisiana. Mountain Press Publishing, Missoula, Montana. 225 p.
STRENG, D. R., J. S. GLITZENSTEIN, AND P. A. HARCOMBE. 1989. Woody seedling dynamics in an east Texas floodplain forest. Ecol. Monogr., 59:177-204.
TANNER, E. V. J., V. KAPOS, AND J. R. HEALY. 1991. Hurricane effects on forest ecosystems in the Caribbean. Biotropica, 23:513-521.
TRAHAN, L. J. 1989. Soil survey of Orleans Parish, Louisiana. U.S. Department of Agriculture, Soil Conservation Service, Washington, D.C. 89 p.
U.S. DEPARTMENT Or COMMERCE. 1969. Hurricane Camille: Preliminary report, http://www.nhc.noaa.gov/archive/storm_wallets/aflantic/af11969- prelim/camille/.
U.S. FISH AND WILDLIFE SERVICE. 2009. Bayou Sauvage National Wildlife Refuge Comprehensive Conservation Plan. U.S. Department of Interior, Atlanta, Georgia. 158 p.
WALKER, L. R. 1991. Tree damage and recovery from Hurricane Hugo in Luquillo Experimental Forest, Puerto Rico. Biotropica, 23:379-385.
WALL, D. P. AND S. P. DARWIN. 1999. Vegetation and elevational gradients within a bottomland hardwood forest of southeastern Louisiana. Am. Midl. Nat., 142:17-30.
WHITE, D. A. AND S. A. SKOJAC. 2002. Remnant bottomland forests near the terminus of the Mississippi River in Southeastern Louisiana. Castanea, 67:134-145.
WILLIAMS, K., M. V. MEADS, AND D. A. SAURBREY. 1998. The roles of seedling salt tolerance and resprouting in forest zonation on the west coast of Florida, USA. Am. J. Bot., 85:1745-1752.
YIH, K., D. H. BOUCHER, J. H. VANDERMEER, AND N. ZAMORA. 1991. Recovery of the rain forest of southeastern Nicaragua after destruction by Hurricane Joan. Biotropica, 12:106-113.
YOU, C. AND W. H. PETTY. 1991. Effects of Hurricane Hugo on Manilkara bidentata, a primary tree species in the Luquillo Experimental Forest of Puerto Rico. Biotropica, 23:400-406.
ZAR, J. H. 1999. Biostatistical analysis, 4th ed. Prentice-Hall, Upper Saddle River, New Jersey. 929 p.
SUBMITTED 2 JULY 2010
ACCEPTED 5 JANUARY 2012
JEROME J. HOWARD
Program in Conservation Biology, Department of Biological Sciences, University of New Orleans, New Orleans, Louisiana 70148
Table 1.--Damage to Gulf Coast hardwood forests from hurricanes. Damage includes major limb or canopy loss, snapping off, and uprooting. To facilitate comparison, maximum sustained wind velocities measured at weather stations in the vicinity of study sites are reported Sust. wind Storm Location Forest type (Km x [hr.sup.-1]) Camille (1969) MS Bottomland 240 (1) Camille (1969) MS Mixed mesic -- Kate (1985) FL Mixed 115 (2) Bonnie (1986) TX Mixed mesic 130 (3) Andrew (1992) LA Bottomland 148 (4) Katrina (2005) LA/MS Bottomland 195 (5) Rita (2005) TX Mixed mesic 130 (6) Rita (2005) TX Bottomland 130 (6) Percent of individuals Storm Damaged Dead Source Camille (1969) 70 -- Gunter and Eleuterius, 1973 Camille (1969) 14 -- Gunter and Eleuterius, 1973 Kate (1985) 41 7 Batista and Platt, 2003 Bonnie (1986) 4.5 -- Harcombe et al., 2002 Andrew (1992) 21.8 1.2 Keeland and Gorham, 2009 Katrina (2005) -- 20.5 Chapman et al., 2008 Rita (2005) 31 -- Harcombe et al., 2009 Rita (2005) 22 -- Harcombe et al., 2009 (1) U.S. Department of Commerce, 1969 (2) NOAA, 2002a (3) NOAA, 2002b (4) Rappaport, 2005 (5) Knabb et al., 2006b (6) Knabb et al., 2006a Table 2.--Mean density and basal area of bottomland hardwood forest tree species at Bayou Sauvage National Wildlife Refuge during Oct. 2004, before Hurricane Katrina. Basal area is total area of all individuals censused ([m.sup.2] x [ha.sup.-1]). Rden = relative density; Rdom = relative basal area; Rfreq = relative frequency; IV = Rden + Rdom + Rfreq Density (stems x [ha.sup.-1] Species >10 cm DBH 1-10 cm DBH Acer rubrum L. 15.9 [+ or -] 30.5 135.3 [+ or -] 328 Celtis laevigata Willd. 132.6 [+ or -] 145.2 23.1 [+ or -] 43.1 Diospyros virginiana L. 15.9 [+ or -] 30.5 55.7 [+ or -] 98.2 Ilex decidua Walt. 0 336.9 [+ or -] 682 Ilex vomitoria Ait. 0 15.9 [+ or -] 27.6 Liquidambar styraciflua L. 34.5 [+ or -] 44.0 13.3 [+ or -] 24.2 Quercus nigra L. 13.3 [+ or -] 24.2 0 Quercus virginiana P. Mill. 47.7 [+ or -] 70.6 0 Sabal minor (Jacq.) Pers. -- 2835 [+ or -] 1591 Triadica sebifera (L.) 190.9 [+ or -] 249.6 1055.7 [+ or -] 990 Mmus spp. 47.7 [+ or -] 51.1 141.8 [+ or -] 197.3 Total 494.5 [+ or -] 180.0 4595 [+ or -] 1347 Shannon diversity (H') 0.729 0.499 (0.538 *) Species Basal area Rden Rdom Rfreq Acer rubrum L. 0.56 [+ or -] 0.001 2.6 3.2 41.7 Celtis laevigata Willd. 9.71 [+ or -] 0.08 3.3 32.0 66.7 Diospyros virginiana L. 0.36 [+ or -] 0.01 1.2 1.3 41.7 Ilex decidua Walt. 1.05 [+ or -] 0.02 9.0 4.2 50.0 Ilex vomitoria Ait. 0.06 [+ or -] 0.01 0.6 0.2 33.3 Liquidambar styraciflua L. 0.94 [+ or -] 0.01 1.1 3.5 50.0 Quercus nigra L. 2.20 [+ or -] 0.04 0.3 8.2 25.0 Quercus virginiana P. Mill. 4.19 [+ or -] 0.06 0.9 13.9 41.7 Sabal minor (Jacq.) Pers. -- 53.41 * 100.0 Triadica sebifera (L.) 5.39 [+ or -] 0.06 23.78 20.36 100.0 Ulmus spp. 2.00 [+ or -] 0.03 3.84 7.54 83.3 Total 26.47 [+ or -] 0.07 * Shannon diversity (H') Species IV Acer rubrum L. 47.4 Celtis laevigata Willd. 102.0 Diospyros virginiana L. 44.2 Ilex decidua Walt. 63.2 Ilex vomitoria Ait. 34.1 Liquidambar styraciflua L. 54.6 Quercus nigra L. 33.5 Quercus virginiana P. Mill. 56.4 Sabal minor (Jacq.) Pers. * Triadica sebifera (L.) 144.2 Ulmus spp. 94.7 Total Shannon diversity (H') * Sabal minor omitted from this calculation Table 3.--Sources of mortality of bottomland hardwood forest tree species at Bayou Sauvage National Wildlife Refuge during Hurricane Katrina. Individuals Species Before After Mortality (%) Acer rubrum 58 1 98.2 Celtis laevigata 58 9 84.4 Diospyros virginiana 27 0 100 Ilex decidua 127 16 87.4 Ilex vomitoria 6 0 100 Liquidambar styraciflua 18 2 88.9 Quercus nigra 5 1 80.0 Quercus virginiana 18 12 33.3 Sabal minor 1069 393 63.2 Triadica sebifera 470 167 64.4 Ulmus spp. 67 11 83.6 Total 1923 612 68.2 Source of mortality Species Flooding Windthrow Acer rubrum 57 0 Celtis laevigata 44 5 Diospyros virginiana 27 0 Ilex decidua 111 0 Ilex vomitoria 6 0 Liquidambar styraciflua 16 0 Quercus nigra 4 0 Quercus virginiana 6 0 Sabal minor 675 0 Triadica sebifera 159 8 Ulmus spp. 54 2 Total 1296 15 TABLE 4.--Mean density and basal area of bottomland hardwood forest tree species at Bayou Sauvage National Wildlife Refuge during Sept. 2006, 1 y after Hurricane Katrina. Basal area is total area of all individuals censused ([m.sup.2] x [ha.sup.-1]). Rden = relative density; Rdom = relative basal area; Rfreq = relative frequency; IV = Rden + Rdom + Rfreq Density (stems x [ha.sup.-1] Species > 10 cm DBH 1-10 cm DBH Acer rubrum 0 2.6 [+ or -] 0.0 Celtis laevigata 23.9 [+ or -] 43.3 0 Diospryros virginiana 0 0 Ilex decidua 0 42.4 [+ or -] 105 Ilex vomitoria 0 0 Liquidambar styraciflua 5.8 [+ or -] 18.0 0 Quercus nigra 2.7 [+ or -] 0.0 0 Quercus virginiana 31.8 [+ or -] 59.6 0 Sabal minor -- 1042 [+ or -] 1795 Triadica sebafera 47.7 [+ or -] 105.1 374.0 [+ or -] 568.9 Ulmus spp. 0 29.2 [+ or -] 79.5 Total 114 [+ or -] 105.9 1475 [+ or -] 1708 Shannon diversity (H') 0.558 0.345 (0.245 *) Species Basal area Rden Rdom Rfreq Acer rubrum 0.002 [+ or -] 0.0001 0.1 0.01 8.3 Celtis laevigata 1.97 [+ or -] 0.04 3.1 13.5 25.0 Diospryros virginiana 0 [+ or -] 0 0.0 0.0 0.0 Ilex decidua 0.07 [+ or -] 0.002 4.7 1.9 25.0 Ilex vomitoria 0 [+ or -] 0 0.0 0.0 0.0 Liquidambar styraciflua 0.20 [+ or -] 0.07 0.3 1.3 8.3 Quercus nigra 0.73 [+ or -] 0.02 0.2 3.9 8.3 Quercus virginiana 3.24 [+ or -] 0.06 2.0 28.7 33.3 Sabal minor -- 58.5 * 100.0 Triadica sebafera 1.42 [+ or -] 0.03 29.8 33.5 66.7 Ulmus spp. 0.07 [+ or -] 0.002 1.4 0.5 16.7 Total 7.70 [+ or -] 0.07 * Shannon diversity (H') Species IV Acer rubrum 8.4 Celtis laevigata 41.6 Diospryros virginiana 0.0 Ilex decidua 31.6 Ilex vomitoria 0.0 Liquidambar styraciflua 9.9 Quercus nigra 12.4 Quercus virginiana 64.0 Sabal minor Triadica sebafera 130.0 Ulmus spp. 18.6 Total Shannon diversity (H') * Sabal minor omitted from this calculation
|Printer friendly Cite/link Email Feedback|
|Author:||Howard, Jerome J.|
|Publication:||The American Midland Naturalist|
|Date:||Jul 1, 2012|
|Previous Article:||Throughfall chemistry and soil nutrient effects of the invasive shrub Lonicera maackii in deciduous forests.|
|Next Article:||The distribution of Quercus rubra in the Maumee Lake Plain of Southeastern Michigan.|