Gray pine (Pinus sabiniana): codominant-stem failure in the utility corridor
No California rule names gray pine. The species adds a fork that arrives after the pole stage, long after the tree was inventoried. The pull-test research behind codominant ratings is red maple. The gray pine that started the Butte Fire failed for a different reason, eight months after a crew cut its neighbors.
A gray pine over a 12 kV tap owes the clearance any tree owes at that voltage. What the species adds is a structural condition that shows up after the tree was first written down. A young gray pine carries one straight stem. A mature one forks, and the fork is printed in the published species description rather than discovered by an inspector.
Three questions follow from that, and the clearance distance answers none of them:
when the second stem appears
what a forked union is worth in a failure rating
what can be prescribed once the union is 40 feet up
There is a fourth question, and it is the one the record of actual California failures answers. The gray pine that contacted a conductor and started the Butte Fire did not fail at a fork. It went over in a 7 mph breeze, for a reason a crew had created eight months earlier.
The fork is in the species description
The USDA Forest Service chapter on the species, in Silvics of North America, Agriculture Handbook 654, is credited to Robert F. Powers and published December 1, 1990. On stem form:
The stem form of Gray pine seemingly disregards gravity. Even on steep slopes it may grow nearly perpendicular to the ground. Trees usually maintain a straight, conical form into the pole stage, but mature trees generally are twisted and have multiple forks.
Read the second sentence as a timeline. The pole stage is when a right-of-way inventory usually meets the tree, and it is the stage at which the tree looks least like what it will become. The chapter's explanation of why names the corridor's own geometry:
The poor form of mature trees probably is a genetic trait but may be traced partly to an open-grown nature that exposes the trees to the damaging effects of wind and to the tendency of lateral buds to elongate when the terminal bud remains static
Open-grown is what a cleared easement makes a tree. The same chapter reports early height growth averaging as much as 28 inches a year for the first 8 years, often exceeding 3 feet in specific years, and puts mature trees at 40 to 80 feet tall and 12 to 36 inches through.
Put a date on that. A gray pine entered in a 2006 pre-inspection as a single 30-foot stem, set back 50 feet from a 12 kV tap, is inside the mature height band by the time a 2026 patrol reaches it. The tree has been visited twenty times. The species field still says gray pine and the structure field, if there is one, still says what somebody typed in 2006. The growth assumption behind the cycle governs the clearance number. Nothing in it re-reads the stem.
There is no failure profile for gray pine, and that is itself the finding
The Western Tree Failure Database publishes the species with 50 or more failure reports. As of January 10, 2026 there are 24 of them. Coast live oak leads at 797, Monterey pine follows at 592, and the list runs down through knobcone pine at 62 and blue oak at 54.
Pinus sabiniana is not on it.
Blue oak is gray pine's partner in the Blue Oak-Gray Pine cover type, Society of American Foresters Type 250. The two grow in the same band around the Central Valley, on the same spans. One of them has 54 filed failure reports and the other has fewer than 50. The database explains the gap itself:
Report numbers for species do not provide an assessment of the frequency of failure for the species (i.e., how often a species fails relative to its occurrence in a population of trees). Rather, it is likely the data reflect the relative abundance of a species in the areas from which reports are being received.
The reporters are urban tree care professionals. Gray pine grows on rocky foothill slopes and canyon walls, which is where distribution circuits cross it and where nobody files a failure report. So a program looking for the species-level evidence base that exists for blue gum will not find one here. The only gray pine failure record available is the utility's own outage and patrol data, and whether that data is queryable by species is a decision somebody made when the system of record was configured.
What splits a fork, in the research that exists
Two pull-test studies are the ones usually cited when a codominant union goes in a file. Both were done on red maple, and saying so is the honest first sentence of any use of them.
E. Thomas Smiley harvested 26 red maples at the Bartlett Tree Research Laboratories in Charlotte, North Carolina between June 1999 and July 2001, took 84 codominant unions off them, and winched them apart. Stem diameters ranged from 4.9 to 23.4 cm, which is roughly 2 to 9 inches. Breaking force ran from 64 to 2,363 kg. Published in Journal of Arboriculture 29(2), March 2003.
The result most people carry away is that included bark weakens a union. It does. The size of the effect is the part that matters in a corridor:
Using results from the regression analysis at 10 and 25 cm, the 10 cm stems are almost 20% weaker when bark is present. At 25 cm, included bark stems are only 14% weaker than nonincluded bark unions.
And the conclusion Smiley drew from his own numbers:
Due to the relatively low reduction in breaking strength at larger diameters, all codominant stem junctions should be considered weak. If trees with codominant stems have a target present that could be damaged if failure occurs, remedial treatments should be applied.
An energized conductor is a target. A field practice that records included bark as the trigger and treats a tight clean union as a pass has inverted the paper it rests on.
Edward Gilman's companion study, in Journal of Arboriculture 29(5), September 2003, tested what predicts attachment strength. One sentence from it should change what gets written on the form:
Angle of attachment was not related to strength of branch attachment. Codominant stems were far easier to split apart than branches that were small relative to stem size. Codominant stems broke by separating at the union; branches failed by breaking at or near the collar.
Aspect ratio was related, at an R-squared of 56 percent: a stem is better held when it joins a much larger stem than when it joins one its own size. Angle was not. A note reading "narrow V crotch" therefore records the one variable the study could not tie to strength, and leaves out the two numbers it could.
So a gray pine forking at 22 feet into stems measuring 14 and 13 inches gets recorded as a ratio of 0.93 and a union height of 22 feet, not as a V. Four fields do the whole job:
Field | What goes in it |
Union height | Feet above grade to the point of attachment |
Stem A diameter | Inches, measured above the swell |
Stem B diameter | Inches, measured above the swell |
Bearing of each stem | Which stem leans over the conductor, in degrees or by phase |
Both studies stop at 9-inch stems. A 30-inch gray pine fork sits outside their range, and the extrapolation is yours to make and to defend. That gap is what a rating against a defect list with no thresholds in it is actually spanning, and the utility owns the method it uses to span it.
The pine California documented at a conductor did not fail at a fork
On the afternoon of September 10, 2015, consulting arborist Michael T. Mahoney met CAL FIRE Battalion Chief Christopher Van Cor at a staging location on Butte Mountain Road outside Jackson and walked into the origin area of the Butte Fire. His report, dated February 15, 2016, became Attachment 2 to the CPUC citation. What he found:
Attention was directed to the ongoing inspection of a Gray pine (Pinus sabiniana) observed with the length of its slender stem lying horizontal, just above the ground with the base of the tree still firmly upright and rooted
The base was still rooted. The stem had laid over. The report puts the top of that tree at 43.7 feet and the charring at 41.5 feet, about two feet below the tip, and calls the burn marks characteristic of contact with energized electrical equipment. The fire ignited at 2:26 p.m. on September 9, 2015. In downtown Jackson, winds at that moment were gusting to the south at 12 km/h, about 7 mph.
The mechanism was the stand, not the tree:
Trees that are captured, and that develop within a closed stand, have narrow growth rings and poor stem taper. When the stand is altered and captured interior trees are exposed to open spaces, whether for right-of-way clearance or other reasons, they are prone to failure.
Ring counts told the rest. Two larger pines at the south edge of the stand had been cut, and the outer ring on their stumps put the cutting before the 2015 growing season began. The subject tree carried about 17 annual rings, close to the same age as the stumps. The report's finding is blunt: the tree lacked reaction wood and was unable to stand on its own, and recent tree and line maintenance activities caused the tree to fail.
Two details from that site are worth carrying into any span with a pine stand on it. The first is that the tree was already tall enough to reach the conductor in the 2014 growing season, before the neighbors came out. Height alone was measurable a full cycle before anybody touched the stand.
The second is a different pine, 45 to 50 feet east of the subject tree and 12 feet north of the conductors, standing there missing its top with branches broken, abraded or missing. Evidence of prior contact was in the same span, in a tree nobody had needed an instrument to see.
January 6, 2015 is the date printed on the violation
The Commission's Safety and Enforcement Division dated the violation from a day that is neither the ignition nor the last patrol. Citation E.17-04-001, issued to PG&E on April 25, 2017, states it as one line in the analysis table:
One violation of GO 95, Rule 31.1 from January 6, 2015 to September 9, 2015 (246 days) for not identifying the subject grey pine tree as a hazard or as needing trimming or removal to prevent contact with a 12 kV conductor.
The clock started the day the two neighboring pines were removed. SED's statement of facts says what the removal did:
SED's investigation found that neither PG&E nor its contractors took appropriate steps to remedy the condition and consequences when two grey pine trees in a stand were removed. The appropriate steps were not taken to prevent a remaining grey pine tree from leaning and contacting the 12 kV overhead conductor.
Rule 31.1 is where a tree that meets every clearance distance still reaches the code:
Electrical supply and communication systems shall be designed, constructed, and maintained for their intended use, regard being given to the conditions under which they are to be operated, to enable the furnishing of safe, proper, and adequate service.
The rule carries its own compliance test for everything the order does not spell out: a company complies "in accordance with accepted good practice for the intended use and known local conditions." Accepted good practice is not a table you can look up.
It is what the Commission will decide after the fact, which is why the citation's phrase is worth keeping: the condition was created by the removal. The CPUC's published text notes the rule was revised January 13, 2006 by Decision 05-01-030 and revised January 12, 2012 by Decision 12-01-032.
Two things in the citation's own analysis table should be read by anyone building a removal workflow. PG&E had an annual distribution patrol, used LiDAR and spectral imagery to find hazardous trees in high fire danger areas, and still, in SED's words, "failed to identify the hazardous condition that was created after two other gray pines were removed."
And SED recorded that its incident investigations had found PG&E in violation of Rule 31.1 37 times since 1999. The penalty mechanics behind that citation and what has to be reported when a facility may be linked to an ignition are their own subjects.
The practical consequence is a workflow item, not an opinion. When a unit removes a tree from a stand, the trees left standing have new exposure, and the date of the removal is the date their condition changed. If the only record of that day is the removal unit closing, nothing in the file says the neighbors were looked at.
The second gray pine failure mode arrives in February
The fork is a summer and wind problem. The roots are a winter one, and the Silvics chapter puts it on a specific kind of ground:
However, hardpan soils are common along the margins of California's Great Valley, and this causes trees to have spreading but shallow root systems with weak taproots extending through duripans. Large trees growing on such sites are windthrown easily during the winter if windstorms coincide with waterlogged soil conditions.
Two conditions have to coincide: wind and saturated soil. That is a January or February storm on a valley-margin flat, not a red flag warning in September. The chapter adds that gray pine is particularly susceptible to damage by wind and hail generally.
A gray pine on a rocky slope has a taproot. The same species on hardpan spreads shallow roots through a duripan instead, and that difference sits under the ground, out of reach of any patrol. Soil is a mapped attribute. If a circuit crosses the valley margin, the duripan is on a soil survey somebody can pull once and attach to the circuit, and after that the winter storm has a list of spans attached to it.
What can actually be prescribed on a mature fork
UC Cooperative Extension's Pruning Conifers, by Donald R. Hodel and Dennis R. Pittenger, gives the textbook answer:
Double leaders should be thinned to one unless the natural growth habit includes several main branches. Encourage branches with wide angles of attachment and smaller than the trunk from which they arise.
That is a nursery and young-tree instruction, and it is the right one at the pole stage. On a mature pine the cut loses its landing place. The same publication explains why:
These trees may have all their lateral buds in whorls just below the terminal bud (most pines), or lateral buds may be scattered along the shoot.
And the consequence:
If there are no visible latent buds, pruning into old wood will usually result in a stub from which no new growth will arise.
Canary Island pine is named in that publication as the notable exception, with latent buds surviving under the bark on large branches and even the trunk. Gray pine is not named as an exception. So on a 70-foot gray pine forking at 22 feet, taking the stem that leans over the conductor is removing half the tree, and the crown does not grow back into the gap.
The honest prescription set is short: remove the whole tree, remove one stem and accept a permanently one-sided crown, or leave it standing with a rating and a date. Which of those is available is a decision with its own criteria and its own approval path, and a fork 40 feet up is where that decision gets made with the least information.
What a gray pine record has to carry
The CPUC's own citation spells the species two ways on facing pages, gray pine in the violation statement and grey pine in the analysis table. That is harmless in a document read once. In a vegetation system of record it is the difference between a species report that returns the tree and one that does not, which is the same problem as a record that says pine instead of Pinus sabiniana.
Five fields turn a gray pine entry into something a later inspector can act on:
Field | Why it is on the list |
Botanical name, spelled one way | The species is queryable or it is not |
Union height and the two stem diameters | The variables Gilman tied to strength; the angle is not one |
Which stem bears toward the conductor | A fork that splits away from the span is a different tree |
Stand context and the date it last changed | The Butte Fire violation was dated from the day the neighbors were cut |
Soil setting, slope or valley-margin hardpan | Decides whether the tree has a taproot or a winter windthrow risk |
The last two are the ones that are usually missing, and they are the two the documented California failure turned on. A fork is visible from the road on a good day. A stand that was thinned eight months ago is visible only in the work history of the span, and only if somebody thought to write the standing trees into the removal unit before it closed.
Further reading
Eucalyptus and power lines: What a species page looks like when a published failure profile exists for the tree.
What makes a tree a hazard tree: The defect list a codominant union sits on, and the rating window it has to be read against.
Level 1 and Level 2 in the field: Whether anybody walks the uphill side of the pine, and which clause of the scope of work bought that.
Frequently asked questions
Does any California rule set a different clearance for gray pine?
No rule names the species. GO 95 Table 1 is written in voltages and distances and PRC 4293 in voltages and radial feet, and a species reaches either number only through the utility's own specification. The rule that reached the gray pine in the Butte Fire was neither of those. CPUC citation E.17-04-001, issued April 25, 2017, cited PG&E under GO 95 Rule 31.1 for failing to maintain its 12 kV overhead conductors safely and properly, and dated the violation from January 6, 2015, the day two neighboring pines came out.
Is included bark the thing that makes a codominant union a finding?
The study most often cited for that says the opposite at working diameters. E. Thomas Smiley pulled 84 codominant unions apart in red maple and found the included-bark unions almost 20 percent weaker at 10 cm and only 14 percent weaker at 25 cm. His conclusion: "Due to the relatively low reduction in breaking strength at larger diameters, all codominant stem junctions should be considered weak. If trees with codominant stems have a target present that could be damaged if failure occurs, remedial treatments should be applied." An energized conductor is a target. Recording included bark as the trigger and a clean union as a pass reverses the paper's finding.
Should the crotch angle go in the record?
It can, but nothing measured connects it to strength. Edward Gilman's 2003 pull tests on Acer rubrum state it flatly: "Angle of attachment was not related to strength of branch attachment." The variable that was related is the aspect ratio, the diameter of the stem relative to the stem it joins. A field note reading narrow V crotch records the thing the study could not tie to strength and omits the thing it could. Two diameters and a height take the same ten seconds.
Can a forked gray pine be pruned back to one leader?
On a young tree, yes. UC Cooperative Extension's Pruning Conifers says to thin double leaders to one and to encourage branches with wide angles of attachment and smaller than the trunk from which they arise. On a mature pine the cut loses its landing place. Most pines carry their lateral buds in whorls just below the terminal bud, and the same publication warns that where there are no visible latent buds, pruning into old wood will usually result in a stub from which no new growth will arise. Removing one of two 60-foot codominant stems is half the tree, and the hole does not fill back in.
Is there a published failure profile for gray pine the way there is for blue gum?
No. The Western Tree Failure Database lists 24 species with 50 or more failure reports as of January 10, 2026, running from coast live oak at 797 down to blue oak at 54. Pinus sabiniana is not among them. The database itself cautions that report counts likely reflect the relative abundance of a species in the areas reports come from, rather than how often it fails. Gray pine grows in the foothills rather than on the street, and the reporters are urban tree care professionals. For this species the only failure record available to a program is its own outage and patrol data.
