Impacts from a Major Ice Storm: Street-Tree Damage in Rochester, New York

1991 storm

According to the U.S. Weather Bureau, the frontal system responsible for this 50-100 year icing event deposited at least 0.8 inch of glaze on tree limbs, power lines, and telephone transmission towers in an area spanning 7,622 mi² (10). Several transmission towers were toppled by the weight of the ice, resulting in a loss of power to 300,000 customers (22). Mroz (20) examined meteorological data for Monroe County for the 1905-91 period and determined that the 1991 storm was the most severe on record for several reasons. The 1991 storm had the most freezing rain (1.5 inches), ranked second in recorded glaze accumulation (0.75-1.0 inch), and had the fifth longest icing event (22.6 hours). During the height of the storm (which occurred at 3:51 am, March 4, 1991), the Rochester weather office reported northeast winds gusting to 24 miles/hour.

Weather conditions after an ice storm, such as persistent sub-freezing temperatures and high winds, can significantly contribute to tree and infrastructure damage, and prolong repair periods. On March 7, three days after the icing event, winds gusted to 51 miles/hour, producing additional damage to trees, poles, and lines. Winds of this magnitude have never been recorded in this region so shortly after an ice storm. Mroz (20) concludes, based on his review of the record and the frequency and intensity of other ice storms, that this storm should be classified as a rare and extreme event.

Glaze damage to trees

Most detailed studies on the mechanics of tree damage from ice storms have focused on natural stands, not urban trees. However, many general observations from these works can be applied to street and lawn trees. First, it is important to recognize that there are different sources of potential damage to a given tree. Failure can be the direct result of glaze, indicating susceptibility due to habit or age, or it may occur indirectly from falling trees ortheir parts (2). In the latter case, damage is most often to smaller trees but no correlation can be drawn between ice accumulation and tree failure; the resulting damage is coincidental. The extent of direct glaze damage to trees has been attributed to the following characteristics: crown size, shape, position and symmetry; trunk diameter; branching pattern and weak branch crotches; and age and degree of decay (3,12,22). Each of these is discussed below.

In general, trees with cylindrical or long, symmetrical crowns suffer less damage from glaze accumulation than those with conical, short, or one-sided crowns (6). In forested stands, needleleaved trees suffer damage directly from ice rather than secondarily from falling trees. They also experience more severe damage (i.e., uprooted or >50% stem and branch loss) than slight damage (i.e., bent over or 50% stem and branch loss) (2). Contrary to these reports, only 22 (5.3%) of the 413 conifers growing on Rochester streets at the time of the storm were severely damaged.

Broad-leaved trees, however, typically have more secondary than direct damage; large trees are severely damaged and small trees are slightly damaged (2). Although it cannot be determined with the available data, some of the tree damage in Rochester may have been caused by indirect effects, perhaps creating the impression that certain species are more susceptible to ice-loading than, in fact, they are.

Downs (9) noted that damage during the 1936 storm in Pennsylvania and New York was more serious in old- and second-growth forests. Presumably, increased damage with age is a function of increased crown size and internal decay, and decreased flexibility of limbs and bole (24). Similar observations were made regarding the 1991 storm in New York (22). Forests with larger diameter trees suffered more damage (30%) due to large, unsound limbs, while those with smaller diameter trees had approximately 15% damage. Pioneer trees, particularly those with an excurrent growth form, suffer more damage than equilibrium species (26).

Most of the storm damage to Rochester’s street trees occurred in intermediate (10-18 inches) and large (19 inches or more) size classes, suggesting that they may have had unsound limbs, or a combination of decreased flexibility and poor growth form (e.g., the acute branch angles typical of callery pear and green ash). Thirty-nine percent of the total Rochester street-tree population was 9 inches or less; only 14.6% of the R1s and R2s, or approximately 3.0% of the total population, was 9 inches or less.

Largertrees suffer mainly from loss of branches and breakage of main stems within the crown. Pole-size trees have more juvenile wood and are often broken below the crown, and small poles and saplings are often badly bent over (1). These different responses to glaze accumulation are most apparent when comparing different aged populations of the same species. Populations dominated by larger, older individuals suffer mostly severe crown and bole damage, whereas those dominated by small stems and/or subcanopy individuals tend to bend over rather than snap off or become uprooted (2).

Initial glaze damage to trees seems to be confined to faulty limbs and functions as a natural pruning process (15). Such branch and limb breakage is not necessarily fatal to trees in a managed setting. According to Croxton (6), if wounds are properly cared for, the effect of glaze damage, in some situations, is similar to that of heavy pruning during a routine maintenance cycle. However, when accumulations exceed 0.5 inch, healthy young branches and well-formed trees may also break. This is evidenced by the breadth of species and number of trees damaged by the 0.8 inch of glaze in Rochester. Ice accumulations have been noted to increase branch weight by as much as 15-30 times (1, 12); the total ice load a tree might bear depends on its size and branching, and on ice thickness. A 50-foot evergreen, for instance, may bear as much as 50 tons of glaze (23).

Attempts to correlate wood strength via comparisons of different species’ modulus of rupture, modulus of elasticity, and specific gravity to glaze susceptibility have met with little success (12,15,18). Wood strength significance is intuitively obvious, but glaze susceptibility is not the result of any single factor and cannot, therefore, be associated with this property alone (6, 15). A given species’ wood characteristics and strength vary, depending on the individual tree’s age, degree of decay present, and local growing conditions, such as available moisture and space, and soil type (3, 15). For example, a large diameter, open-grown oak more than a century old may suffer extensive ice-storm damage, whereas the same species, 50 years old, may suffer little to no damage. Hauer et al. (12) maintain that the strength of wood is less important in ice-accumulation tolerance than the tree’s ability, based on its growth form, to accumulate ice until the bearing capacity of its branches has been surpassed.

Branching pattern, specifically branch angle and the amount of fine branches, affects resistance to ice damage (5). Trees with horizontal branching have a higher susceptibility to damage because this orientation favors glaze accumulation through greater exposed surface area. In Rochester, this was evidenced by damage to honeylocust, London planetree, Japanese pagoda, and Norway maple (Pleninger, A.G., Rochester City Forester, personal observation, 1992). Conversely, branches growing upward at an acute angle are oriented perpendicular to the exposed surface of the crown, resulting in lower overall glaze exposure. Species with opposite branching patterns tend to have wide branch and twig angles, and therefore suffer more storm damage (3). Contradictory to these data, the maples in Rochester fared relatively well, except for larger size classes. Structurally weak crotches, especially those with included bark, also increase damage in an ice storm (4). Storm damage to green ash, callery pear, and red maple in Rochester has been attributed, in part, to weak crotches and included bark (Pleninger, A.G., Rochester City Forester, personal observation, 1992).

Numerous small twigs and branches increase total surface area, resulting in greater ice accumulation and damage (15). In addition to causing differences in the amount of accumulated glaze, branching pattern and twig size also influence the type of damage. Small twigs are relatively flexible and tend to bend under added weight. Glaze will remain on these twigs, concentrating weight onto larger, less flexible branches that may break under stress. This pattern of damage was observed in green ash, honeylocust, and Norway maple in Rochester (Pleninger, A.G., Rochester City Forester, personal observation, 1992). By contrast, relatively large twigs accumulate less glaze per unit diameter. However, these are less flexible and snap off distally to a greater extent than those of finely branching species (3).

Species susceptibility and glaze damage in Rochester

Ice-storm tolerance ratings for tree species vary for different geographic locations and from natural stands to urban settings (2,3,12). This is due to differences in local topography and climate, as well as site conditions, plant age and structural integrity, and past maintenance practices (2). Richards (21) notes that “within New York, there are major differences in species adaptation among climate and soil regions.”

Examples of variable tolerance exist for many of the Rochester storm-damaged species. Seischab etal. (22) rated species susceptibility to ice damage in western New York based on observations in natural stands following the 1991 storm. They found that ash as a group had minimal damage; green ash had 10% and white ash (F. americana) had 17% crown damage. By contrast, Bruederle and Stearns (3) rated white ash as highly susceptible in southern Wisconsin. In the urban settings of Urbana, Illinois (12), and Rochester, New York, green ash sustained severe damage. Fifty-one percent of the green ash population in Rochester was lost to the 1991 ice storm. These differences between urban and natural trees may be due to the less-protected, open grown nature of shade trees (1,26).

Species variety might contribute to ice tolerance as well. For example, Marshall’s seedless ash, a variety of green ash commonly planted in public settings, including Rochester, has a more upright form with more acute branch angles, a dense branching pattern, and weaker crotches than forest trees of the same species. This difference in habit results in different responses to glaze accumulation. Marshall’s seedless ash is tolerant of harsh growing conditions in urban areas, but does not fare as well in ice storms as native forest trees.

Conflicting reports also exist for some of the maple species found in Rochester. Extensive ice damage to silver (central Illinois (6); PA to Boston (7); Syracuse, NY (16)), red (North Carolina (1,7)), and Norway (Syracuse, NY (16)) maples grown as public trees has been observed. However, data supporting the glaze tolerance of these trees are available as well; some are presented below.

Silver maple sustained only moderate damage in Urbana, Illinois (12). The authors observed that this species typically breaks at upper- or midbranch positions when loaded with ice, and suggest that damage was low because these trees had been regularly pruned since 1976. Although 35.9% of the silver maple population in Rochester suffered severe storm damage, most (82.3%) affected trees were 19 inches dbh or more. The damaged trees in Urbana were mostly greater than 24 inches dbh. Differences in the amount of damage to silver maples in Urbana and Rochester might be attributable to past pruning practices.

Similarly, red maple sustained little damage in Urbana, those that were impacted were greater than 24 inches dbh (12). In Rochester, 16.8% of the red maple population was severely damaged by glaze. More than 62.0% of the storm-damaged red maples were 19 inches or more.

A clear demonstration of variable tolerance and interpretation of results is that Norway maple’s success (7) and susceptibility (16) have been attributed to its crown compactness and branching pattern. Norway maple in central Illinois had no damage (6) and those in Urbana sustained little damage (12). Likewise, Norway was the most resistant maple following the 1940 storm which stretched from Pennsylvania to Boston (7). In Rochester, 20.1 % of the Norway maple population was lost to glaze damage; however, most of these were generally in poor condition before the storm (Pleninger, A.G., Rochester City Forester, personal observation, 1992).

Sycamore (Platanusoccidentalus) is used here to review ice susceptibility of London planetree because these species are closely related and structurally similar (both have horizontal branching resulting in high surface area (8)), and response to ice loading is assumed to be comparable. In Urbana, sycamores sustained severe damage (12). The authors suggest that this was the result of anthracnose increasing surface area and elevating susceptibility to ice damage, and is not a direct reflection of this species’ ice tolerance (12). Twenty-five percent of the Rochester London planetree population was lost to glaze damage, and the city forester reports similar increases in surface area due to past episodes of anthracnose (Pleninger, A.G., Rochester City Forester, personal observation, 1992).

Hauer et al. (12) maintain that Bradford pear, a variety of callery pear, had extensive damage in Urbana because its narrow branch crotches, often with included bark, form weak attachments. They suggest that the Aristocrat pear performed better (none was damaged) because its branches are more horizontal with a wider crotch angle. Only one representative of this cultivar was present in Rochester at the time of the storm, it suffered minimal glaze damage. Approximately 24% of the Bradford pear in Rochester was storm damaged.

Littleleaf linden suffered minimal damage in Urbana (12); only 13.4% of the population was lost in Rochester. Storm damage to honeylocust in Urbana was attributed to structural imperfections, disease, or insects (12); 18.3% of the species’ population was lost in Rochester. No glaze storm data from other cities could be found for Japanese pagoda, however, 63.1% were lost in Rochester.

Management Implications

Specific recommendations regarding the primary storm-damaged trees in Rochester are outlined below, with some observations on their past performance provided by the city’s forester.

Four species—silver maple, green ash, callery (Bradford) pear, and honeylocust—that were severely damaged in Rochester were also severely damaged in Urbana, Illinois (12). Possible reasons for the failure of silver maple have already been discussed; some of its advantages will be outlined below. The remaining three trees seem to consistently fare poorly in ice storms and may be inappropriate in areas that experience frequent or severe glaze events. Other species that were extensively damaged in Rochester include Japanese pagoda, London planetree, Norway and red maples, and littleleaf linden. These had less damage (RSD values less than or equal to one) but still lost 15% to 20% of their populations.

The disadvantages of green ash (e.g., acute branch angles, dense secondary branching, included bark, weak crotches) suggest that its use should be reduced. However, it is very urban tolerant and newer cultivars have better branching patterns than the Marshall’s seedless variety (11).

Strong evidence for success or failure does not exist for callery pear and honeylocust indicating that their continued use in Rochester should be more closely evaluated. Bradford callery pear has a late leaf drop (making it susceptible to damage during early, wet snowfalls), and a poor branching pattern, weak crotches, and included bark. Other cultivars may perform better.

The city forester in Rochester favors continued use of honeylocust despite its susceptibility to glaze damage. It is very tolerant of urban conditions, and performs well on heavily salted arterials, in a variety of soils, and in limited space. Its overall success in non-glaze situations warrants risking damage from rare icing events like the 1991 storm.

Silver and red maples should continue to be planted in Rochester. Much of their individual populations are 19 inches or more (73.7% and 34.2%, respectively), demonstrating their potential longevity. Most damage to these species was sustained by larger trees, suggesting some glaze tolerance when they are younger. Based on stormdamage data alone, silver maple, in particular, is a strong candidate for street-tree planting in Rochester. It is tolerant of urban conditions, very long lived, can weather ice storms if pruned regularly, and was found to be healthier than sugar and Norway maples in Poughkeepsie, Syracuse, and Rochester, New York (25). However, 2- to 3-inch limbs often break off of larger trees, creating wounds that are prone to decay. Maintaining this species at a level that minimizes limb breakage requires a 2- to 3-year pruning cycle, which is difficult to establish in a city like Rochester with a large street-tree population. Red maple requires acid soils and good training when young, but is a good residential street tree in Rochester where site and space requirements are met.

Japanese pagoda should not be planted because 63.1% of its population was lost to glaze. Despite this species’ tolerance of urban conditions, it has a variety of shortcomings in addition to its susceptibility to ice damage. It produces a lot of litter, especially petalsduringitsextensive flowering period, and fruits persist year-round. In addition, its flowers attract bees, leading to many complaints from residents.

Norway maple should be used less (it currently comprises 27.7% of the total population) since it represented more than 27.2% of the stormdamaged trees. Its use should not be eliminated completely, however, because it can endure urban conditions, as evidenced by the many older trees in the Rochester population (32.1% are 19 inches or more). It performs well in a variety of soils and in Rochester’s climate, but requires good rooting space and will decline rapidly after sustaining severe injuries. Several varieties, Schwedler, Crimson King, and columnar, fared well in the 1991 storm, suggesting that their use as street trees may be more appropriate.

Littleleaf linden has been a good residential street tree in Rochester and will continue to be planted. London planetree has good urban tolerance and longevity, compartmentalizes well, and can survive significant crown loss. Both of these species had a smaller proportion of storm loss than their individual populations.