ArbNation Join free

Why trees fail in storms: what University of Florida hurricane studies report

University of Florida researchers measured urban trees after 10 hurricanes, from Andrew in 1992 to Katrina and Rita in 2005. Trees fell by uprooting, breaking or leaning, and wind speed, soil, age, decay and pruning all shaped the losses.

Checked against the sources at the bottom of this page on October 15, 2026. Rules, fees and pay change: the source has the last word.

University of Florida Institute of Food and Agricultural Sciences (UF/IFAS) researchers have measured yard and street trees after hurricanes since 1992. Their fact sheets FR173 and FR174 and a two-part journal paper in Arboriculture and Urban Forestry carry the findings below. All figures are theirs.

How the studies counted failure

The team measured standing, leaning and fallen trees in neighbourhoods at landfall. Trees that were standing counted as survivors. Trees that had fallen, or were leaning at less than 45 degrees, counted as not surviving.

Fallen trees were sorted as broken at the main stem or uprooted. FR173 names the three damage types as uproot, break and lean. For standing trees, the team also scored branch loss and leaf loss. A tree that lost 50% or more of its branches was declared dead in the recalculated figures.

FR174 reports that the team sampled 100 neighbourhoods and 18,200 trees in the coastal plain work. FR173 says the full study covered over 150 urban tree species across 10 hurricanes. FR174 counts over eighty species across nine hurricanes for the coastal plain alone.

Wind speed and urban forest loss

Loss rose with wind speed. FR173 puts the share of the urban forest lost at 11% in Erin, 21% in Rita, 23% in Katrina and 38% in Andrew. The tropical paper adds 13% to 18% for three more storms.

Urban forest loss by hurricane
HurricaneLossReport
Erin11%FR173
Georges13%Part II paper
Jeanne16%Part II paper
Charley18%Part II paper
Rita21%FR173
Katrina23%FR173
Andrew38%FR173

FR173 adds that wind alone does not decide the outcome. Rain and the speed the storm moves through also count, along with species, age, health, structure, soil and canopy density.

The fact sheet compares Pensacola, with a denser canopy of older trees, and Miami, with less canopy and younger trees. Pensacola's trees suffered considerably more damage.

Roots and soil

FR173 calls rooting space probably the most overlooked factor in a healthy urban landscape. In Hurricane Georges in Puerto Rico, survival climbed with the soil area open to roots.

Rooting space and survival, Hurricane Georges
Rooting spaceSurvival
0 to 3 m2 (0 to 39 sq ft)64%
4 to 7 m2 (40 to 75 sq ft)73%
Over 7 m2 (over 75 sq ft)91%

The tropical paper agrees: trees with more than 7 m2 of rooting space had the lowest branch loss and the best survival in Georges.

FR173 also reports that trees in shallow soil blow over more often than deeply rooted trees. In Miami-Dade County the soil is no more than one foot deep. A slow, wet storm leaves more water in the soil and less friction between roots and soil.

Cut roots are another path to failure. FR173 shows roots severed during sidewalk and curb repair. Roots that extend out from a tree in all directions stabilize it against wind throw. The closest distance to cut roots is five times the trunk diameter.

Age, size, decay and structure

Older and larger trees lost more branches. In FR174, trees of 100 to 200 cm diameter lost 30% of their branches, and those under 20 cm lost 12%. FR173 links age to insects, disease and stiffer wood.

Decay is named as a major cause of tree failure. Cracks, seams, butt swell, dead branch stubs and old wounds suggest internal decay. Mushrooms at the base of the trunk can point to root decay from Armillaria or other fungi.

FR173 also reports on structure. Codominant stems, two or more stems of equal size from one point, may form included bark, a weak union very prone to breakage. For more on these defects see codominant stems and included bark and root damage and root plate lifting.

Pruning history

FR173 states that topping and removing large branches make trees more open to wind failure, and that old large pruning cuts can let in decay fungi. After Hurricane Andrew, a study of seven broad-leaved species found 73% survival for pruned trees and 47% for unpruned trees. The tropical paper repeats that reanalysis. Storm-season pruning is covered in pruning after storms.

Species and groups

Species differed. FR174 ranks laurel oak below live oak and sand live oak across four Panhandle hurricanes. After Ivan, its recalculated survival ran sand live oak 98%, live oak 81%, laurel oak 66%, water oak 65% and southern red oak 50%.

Palms often did better as a group, because they carry one terminal bud. FR173 notes that queen and Washington palms did poorly in south Florida. Pines are a separate case: three months after Charley, 27% of the standing south Florida slash pines and 48% of the standing longleaf pines had died. FR173 says pines can die over 6 months to 2 years.

Groups of five or more trees, each within 10 feet of another and not in a row, survived better. FR173 gives 80% against 70% in Ivan and 88% against 78% in Jeanne.

The tropical paper reports that, of the trees that fell in Jeanne, Charley and Georges, 3%, 4% and 11% damaged property.

Where the studies differ

On leaf loss, the sources split. FR173 and FR174 report that trees losing leaves in Ivan survived better, and FR173 calls leaf loss a possible strategy. The tropical paper finds that leaf loss had no relationship with survival in Jeanne, Charley and Georges.

On native species, FR174 found natives and exotics survived alike in the coastal plain (73% against 77%), while the tropical paper found natives did better in Jeanne and Charley but not in Georges. FR173 also cautions that no species is completely windproof.

Sources