The California Tree Failure Report Program: An Overview

Materials and Logistics

A failure report form was developed (Fig. 1) to enable participating arborists to collect detailed information about individual tree failures. The form is a comprehensive checklist which serves as a permanent record of each occurrence. In addition to basic items such as tree name, age, size, and location, six general categories of information are requested: 1) details of failure, 2) structural defects, 3) decay or injury, 4) maintenance history, 5) soil and root conditions, and 6) weather. Statistical methods can be applied to data collected for each of these categories so that the type of failure and circumstances which led to the failure can be characterized.

• Download figure
• Open in new tab
• Download powerpoint

The form is designed so that up to three contributing factors within any category can be reported, and these factors can be prioritized in order of importance. For example, under Tree Structural Defects, if a dense crown, onesidedness, and girdling roots contributed to a blowover, then numbers 3, 5, and 12 would be entered in the spaces to the left of the defects list. If it appeared that girdling roots were the most important contributing factor, followed by onesidedness, and dense crown, then numbers would be entered in the order 12, 5, and 3, from top to bottom. If a category does not apply or other circumstances contributed to the failure, then 13 (none) or 14 (other) would be entered. This approach accommodates a database that allows for meaningful statistical comparisons.

The reports are intended to be detailed and do require some time for completion. Once an arborist is familiar with the format and items in each category, however, reports can be completed in 10 to 15 minutes, or less. We consider it necessary to include any possible contributing factors in the checklist so that we do not limit the database.

Many of the common factors involved in failures are included in the checklist, but we recognize that a checklist does not always allow for a complete analysis of a failure. For example, if a tree uproots because an adjacent tree has restricted its root development (anchorage) on one side, then the failure would not be adequately described by the checklist. Space is provided on the second page for a short narrative describing the failure (Fig. 2). Information regarding property damage, personal injuries, costs for clean-up, and additional comments is also requested on page 2. This is an opportunity to report information that is not associated with every failure, but is of interest when evaluating failure impacts.

• Download figure
• Open in new tab
• Download powerpoint

Cooperators

A network of CTFRP cooperators was established to report on failures in their areas. Invitations were extended to qualified tree care professionals to participate in the program. Each was supplied with forms and stamped return-envelopes. To assure quality control in the information reported, all cooperators attended a CTFRP training class. They were instructed on how the form was to be completed, and what was meant by certain terms, such as “lion-tailing” or “fungal sporophores.” It was vital that a “common thinking” was established among cooperators in regard to the information requested. Cooperators were advised that incomplete or inaccurate information would seriously diminish the value of the program. Completed reports were not accepted from individuals who had not been trained.

Completed reports were sent to the Environmental Horticulture Department at the University of California, Davis, where a database program has been designed using dBase IV (Ashton-Tate) to facilitate entry and analysis of CTFRP data. Reports are received, data are entered for each category on the form, and originals are filed. Photographs, slides, newspaper clippings, or other supportive materials enclosed with the report are also kept on file.

Current Project Status

As of December, 1990, 500 CTFRP reports have been filed, with over a hundred cooperators participating state-wide, in twenty-three counties. Approximately two-thirds of the reports have been sent in from arborists in the greater San Francisco area, where we have held the most training sessions. As we have begun to report findings, interest in the program has increased. There is now substantial interest from the San Diego and greater Los Angeles urban areas. We hope to extend our training workshops to include this group soon.

Some Highlights of the Survey

Three categories of data have early interest to us: 1) the location of the failure on the tree, whether trunk, branch or roots (Location of Failure); 2) structural problems associated with the failure (Tree Structural Defects); and 3) the wind speed at the time of failure—low to no wind; medium wind, 5-25 mph; high wind >25 mph.

For the database as a whole, location of failure was fairly evenly distributed among the trunk, branch, and root categories, with more branch failures (42%) than trunk (30%) or root (28%) failures reported. Root failures occurred mostly (67%) in high winds, while branch failures occurred in all wind conditions. Heavy lateral limbs were by far the most common defect in branch failures (75 of 201 branch failures, or 37%), and were the most often-reported structural defect overall. Certain structural defects were associated frequently with high wind conditions (such as embedded bark in branch crotch, branches at the same point, or multiple stems). Failures associated with heavy lateral limbs were reported in all wind conditions, however.

Trunk branch, and root failures have been reported for at least 62 different tree genera. Of these, reports on five genera have been the most numerous: Quercus (110), Pinus (76), Eucalyptus (70), Ulmus (33), and Cupressus (30). Reports on oaks represent about 20% of the total reports f i led. Since there are many native oaks in our city landscapes, as well as pines and Eucalyptus, these data do not necessarily indicate an inherent propensity to failure for these particular genera. We do not yet have estimates of the relative abundance of genera and number of individuals of a species in survey locations.

The high number of reports for these five genera will allow us to look more closely at them, to compare them to the database as a whole and to see whether we can identify a “failure profile” for a particular genus. It is interesting to note for example that 55% of all failures reported for oaks happened in low or no wind conditions. By contrast, only 15% of all failures reported for Eucalyptus happened during low or no wind. Branch failures were the most frequent type of failure in Eucalyptus (>50%); branch and trunk failures were both fairly common in oaks (about 38% each). These and other data suggest that failure profiles can be assembled, which will provide useful information for understanding tree failure, perhaps helping to reduce tree hazards.

Specific Requests

In addition to the above data summaries, specific requests regarding failure types have been received. For example, one cooperator was interested in failures associated with girdling or kinked roots. At the time, approximately 4% of all failures reported were due to girdling roots. Of these, 53 per cent were pines over 30 feet in height. This is of interest because pines are a much smaller fraction (15%) of the total number of reports. Indications are, therefore, that girdling roots may occur more frequently in pines than other species, and they could be an important factor to consider in stability assessments.

Feedback Mechanisms

Three ways of returning information to cooperators have been established: a newsletter, an annual meeting, and personal requests. The newsletter Breaktime is published in the spring and fall of each year. CTFRP report totals, data summaries, interesting features, and program status reports help cooperators stay informed. It is essential that a close link be established with cooperators, and the newsletter has served this purpose.

In addition to the newsletter, an annual meeting is held during the winter months. This is an opportunity for cooperators to meet one another and discuss issues relevant to tree failures. A full day program includes presentations on subjects such as wood decay or sudden branch drop, as well as data summaries and analysis.

Cooperators also receive information regarding particular species or types of failures via personal requests. As noted previously, a request was received for cases where kinked or girdling roots were noted as structural defects. In another case, an arborist was interested in failures associated with severe heading cuts. Entering the appropriate information into the computer, a printout of all heading-cut related failures was generated. This included information on species, age, size, and conditions at the time of failure. In some cases, there may be little or no data relevant to the request. Lack of information can be useful when assessing the frequency of particular types of failures in certain species. For example, we observe very few branch failures in London plane tree (Platanus acerifolia). In checking the database, we find no reports of this failure in this species. Of course we cannot conclude that P. acerifolia does not have branch failures, but it is an indication that the frequency of failure may be low.

These three feedback mechanisms have all served to establish a vital link in the informational loop from data collection to data application.

Other Benefits

Aside from collecting data to make probability of failure estimates, other benefits have been realized. Certainly, public safety has been improved. Cooperators report that because of their systematic inspection of failed trees, they now look more closely at remaining trees for conditions similar to the failure. Such inspections are the cornerstone of hazard assessment programs and are well recognized to reduce failure potential (5). Tree preservation has also been seen as a benefit. Recognizing and mitigating structural weaknesses results in a longer life for many trees.

Perhaps one of the greatest benefits has been professional improvement. By providing athorough mechanism for inspecting tree failures, cooperators have gained a greater understanding of tree structure and function. Determining how a tree fails helps one understand how it is put together. Repeatedly we hear cooperators say the CTFRP form has motivated them to inspect failures more closely than before. Previously, tree clean-up was the first priority after a failure. In some cases, failures could be ignored completely. Now, cooperators inspect and ask questions about the cause of failures. This invariably leads to a greater understanding of trees and therefore improves professionalism.

Application

The CTFR Program can easily be adapted for use in other states and countries. The form and database program may need to be amended for local conditions, but basic elements of the data collection network would remain the same. Certainly it is important to have as many well qualified cooperators as possible, but a small group of active individuals can accomplish a lot.