Hurricane Katrina Winds Damaged Longleaf Pine Less than Loblolly Pine

Saucier-Susquehanna complex. Slope in this complex varies from 0 to 8%, but it was closer to 1–4% on the study site. Five treatments were established: (1) no cultivation or fertilization; (2) cultivated with no fertilization; (3) cultivated with a single application of 112 kg ha 1 of NPK fertilizer (type unknown); (4) cultivated with a single application of 224 kg ha 1 of NPK fertilizer; and (5) cultivated with a single application of 448 kg ha 1 of NPK fertilizer. Cultivated plots were cleared of all stumps and slash, plowed, and then disked prior to planting. They were then disked three times each season for 3 years to reduce woody competition and then mowed in years 4 and 5. Fertilizer was applied with an agriculture spreader and disked into the soil in May, 1 year after planting. In addition to the cultural treatments, the loblolly, longleaf, and slash pine seedlings for the experiment were grown in a nursery using seed from local source parents that were classified by their measured wood specific gravity. Two groups were used in study, those with high specific gravity (HSG) and low specific gravity (LSG) (Clark and Schmidtling 1988). The experimental design was a randomized complete block with split plots, replicated four times, creating 120 subplots (three species two specific gravity treatments five silvicultural treatments four blocks). Whole plots within replications represent the species treatment and they consisted of 10 (two specific gravity treatments five silvicultural treatments) 100-tree plots. Subplots consisted of 100 1-year-old bare root seedlings that were bar-planted in 3.05 3.05 m spacing in February and March of 1961. No operational thinning has been carried out. Fire records are incomplete prior to 1994, but since January 1994 (age 33 years), the study site has been burned in the dormant season five times: 1994, 1998, 2001, 2002, and 2003. Growth through 25 years (fall 1984) has been reported by Smith and Schmidtling (1970), Schmidtling (1973), and Clark and Schmidtling (1988). Height and dbh (DBH) were measured on all trees in the winter of 1999–2000 (40 years in the field). Mean stocking (stems per hectare), height (m), dbh (cm) and basal area (BA, in m ha ), for loblolly, longleaf and slash pine respectively were: stocking (449, 552, 579); height (19.8, 21.0, 22.1); dbh (23.2, 24.6, 25.8); and BA (21.1, 26.7, 28.7). Hurricane Katrina struck the experiment in Aug. 29, 2005 with sustained winds greater than 145 km hr 1 (US Geologic Survey 2008) with peak gusts up to 225 km hr 1 (Kupfer et al. 2007). In Dec. 2006, all trees in the experiment were measured for dbh and rated with respect to damage from the storm. Trees were rated in three categories: undamaged, downed, or boles snapped. Killed trees include those that were downed or with boles snapped. The dbh just prior to the storm for killed trees was estimated by adjusting individual tree dbh growth by the average annual growth between the 1999 and 2006 measurements. Earlier research demonstrated that the initial phenotypic selections for wood specific gravity resulted in no difference between selection types in specific gravity or tree growth across species and treatments (Schmidtling 1973, Clark and Schmidtling 1988). Because there was no specific gravity treatment effect, the HSG and LSG treatments represent noncontiguous plots. For our experimental design, we combined the two noncontiguous 100-tree specific gravity treatment plots into single 200-tree plots resulting in onehalf the original number of subplots (three species five treatments four blocks 60 plots). In addition, initial analyses of pre-Katrina basal area provided a rationale to create three treatments from the five installed in the field. The fertilization treatments (treatments 3, 4, and 5) were not significantly different from each other and so were combined and named cultivated fertilizer applied (C-F), whereas the remaining treatments were left as they were and named no cultivation (NC, treatment 1) and cultivated no fertilizer applied (C-NF, treatment 2). This resulted in a further reduction in the number of subplots to 36 (three species three treatments four blocks). Analysis of variance was used to test for statistical differences among treatments, species, and their interactions. Comparisons between means were made using Tukey’s multiple range test. Analysis of covariance was used to examine the impact of tree height, plot density, and the diameter/height ratio on species variation in tree mortality. Because tree heights were not measured prior to or following the hurricane, we used tree heights measured in 1999 (6 years before the storm). For these regression analyses, all 120 plots (100 trees each) were used with 0.05.