Recurrent tropical cyclones frequently influence warm temperate forests in many coastal regions. If resistance to high winds (cyclone survival) and resilience in altered environments (post-cyclone recruitment) vary with cyclone intensity, then differences in composition, arboreal structure and dynamics of forests should occur in these forests. We hypothesized that major tropical cyclones should produce large, stratum-dependent effects that result in large environmental changes. We anticipate direct regeneration, with post-cyclone forests resembling pre-cyclone forests in composition, but potentially shifting to long-transient alternate states that do not readily or necessarily return to pre-cyclone states. Using a long-term, plot-based study, we explored direct effects of major Hurricane Katrina (sustained winds >50m/sec), as well as initial post-hurricane changes in overstory and understory trees in oak-dominated bottomland and cypress/tupelo-dominated swamp forests within the Mississippi River delta. Hurricane Katrina caused high overall mortality (14% - swamp forest, 25% - bottomland forest), but mortality was concentrated in overstory strata (22% in swamp forests and 70% in bottomland forests). Almost all overstory oaks were killed, while most overstory cypress and tupelo were damaged, but survived and were reduced to understory status. In both forests, many trees of less abundant overstory species (red maples, sweetgum, elms, ash) survived and constitute a slowly reforming overstory via resprouting of survivors and recruitment of understory trees. Thus, forest overstories appear likely to change in dominant species. Understory trees in both forests had high survival and responded post-cyclone. Nonetheless, one usurper species, dwarf palmetto, with high resistance and high resilience, now forms a dense arboreal layer in the understory stratum, dominating an alternate state with different overstory and closed understory capable of suppressing recruitment and establishment of overstory and understory species. Resistance and resilience of overstory and understory tree species that vary with cyclone intensity might result in prolonged alternate states in cyclone-frequented forests.

The current study utilized one section of the larger transect to examine changes in forest tree vegetation associated with Hurricane Katrina. In 2004, 20 non-adjacent plots, each 100m² (10 x 10 m), were randomly established along several hundred m of the transect extending downslope from the top of the western levee of Bayou des Familles to open freshwater marsh. Ten plots were situated in oak-dominated bottomland forest within the more-upland section of the transect that was not flooded in 2004. Ten other plots were situated in seasonally inundated cypress and tupelo-dominated swamp forest. These swamp forest plots were distributed along the transect from upper areas only briefly flooded (1-2 months) to almost permanently flooded areas with numerous hummocks on which woody vegetation was present.

Sampling was conducted four times. We sampled each plot during the summer (August) of 2004 and 2005 (pre-hurricane). We revisited plots for post-hurricane field observations and plot relocation and remarking, as needed, in October, 2005, when we could access plots. We conducted our post-hurricane sampling in September, 2006. Sampling was resumed in the summer of 2017. During our surveys, we systematically located and examined woody vegetation with stems > 1.5 m tall in each plot. We recorded species and arboreal stratum [understory (< 15m), subcanopy (15-20m), and canopy(>20m)] of each plant on plot maps. Abundance of species was estimated using methodology developed by Peet et al. (1998), modified for only woody vegetation and using nested subplots (1, 10, and 100m²). In all sampling years, voucher specimens of woody plant species were collected and deposited within the Shirley C. Tucker Herbarium housed at Louisiana State University in Baton Rouge, Louisiana. Species were identified based on Weakley and Southeastern Flora team (2022). Prior vegetation surveys conducted within the Preserve also provided useful taxonomic references (White et al. 1983, Michot 1984, Michot and Doyle 1999, Swarzenski et al. 2004, Nolfo-Clements 2006, Urbatsch et al. 2009).

Microsoft Excel