Target species: Quercus douglasii, Quercus wislizeni
Other species: a few observations on Arctostaphylos viscida, Cercocarpus betuloides, Rhamnus crocea, Toxicodendron diversilobum, Yucca whipplei
Life stages: trees
Origin: natural
Situation: nongrazed parkland
Location: Sequoia National Park, Kaweah River drainage, Tulare Co. (580 m elev.)
Overview: This paper reports on the survival of Q. douglasii and Q. wislizeni trees for 2 years following the Ash Mountain fire in June 1987. The fire was described as a moderate to severe ground fire. Fire-caused mortality two years after the fire was low, 11% for Q. wislizeni and 6% for Q. douglasii. The extent of mortality was not evident immediately after the fire. Q. douglasii recovered from crown scorching by epicormic sprouting in the crown, whereas Q. wislizeni recovered by basal sprouting from the root collar. Fire scars developed on about 60% of the Q. douglasii, but many of these were not evident until 2 years after the fire. Fire scars formed on about 40% of the Q. wislizeni.
Methods: According to park records, the study site had not been burned for at least 62 years prior to the 6/87 arson fire that is the basis of the study. The site was last grazed by pack stock (horses and mules) 10 years prior to the burn and was not grazed during the study. Fire intensity was inferred from plant damage and retrospective analysis of weather and fuel conditions. No pre-burn observations were made at the study site; post-burn observations began about 3 months after the fire. The author established four nonrandom transects, selected to represent different topographic positions and fire intensities. Plots were spaced at 30 m intervals along 300 m transects. Although each plot should have consisted of 4 tagged trees (selected using the point-centered quarter method), an additional tree was selected for each tagged dead tree. The additional trees bias the total survival percentages downward and add a bit of confusion to the interpretation of survival numbers. If the dead trees were excluded from the survival calculations, the study could be interpreted as reporting on the post-fire survival of trees that were alive immediately after the fire rather than survival of all trees. It appears that the author included all tagged trees in some but not all of the survival calculations (e.g., for Q. wislizeni, n=29 in Fig. 2 but n=31 in Table 2). The error introduced by the use of different values for is on the order of 1-3%.
After the initial tagging, each tree was assessed twice. Dates of these later assessments are not given, but they presumably occurred in the fall of 1988 and 1989. Eighteen months after the fire (12/88 or 1/89), the author established 42 0.05 ha circular plots along the same transects to survey survival in oaks more than 1.37 m tall. However, only 14 "seedling" oaks (>1.37 cm tall and <5 cm dbh) were located, and most of these were in a single transect, so data on this small size class is not very reliable. Also, midwinter is generally a poor time to assess survival in blue oak because it is fully dormant at that time.
Effect of fire intensity. Based on post-fire tree damage symptoms, the fire was described as a severe ground fire, with some areas rated as a moderate ground fire. Calculated estimates of fireline intensity ranged from 315 to 1072 kW/m at the fire front (head fire) to 17 kW/m at the backing edge of the fire. Tree damage was generally positively correlated with the relative fire intensity in each transect. The transect with the greatest fire intensity (along a ridge) had the highest amount of crown scorch, the greatest incidence of complete topkill (19%) and highest incidence of basal sprouting (32%) among the 4 transects.
Post-fire survival. Apparently 4 trees in the nominal 160 tree sample (2.5%) were dead at the time of the original sampling 3 months post-fire. For the oaks in this sample (ignoring additional trees added to compensate for dead trees) 2 of 146 (1.4%) oaks were immediately killed by the fire. By two years after the fire, 6% of tagged Q. douglasii and 11% of Q. wislizeni trees died as a result of the fire. Mortality rates in fixed-area plots were lower for Q. wislizeni (2%) but similar for Q. douglasii (4%).
Factors related to mortality. In logistic regression models, only percent crown scorch and previous fire scars predicted mortality. Since the site hadn't burned for at least 62 years, the presence of any discernable fire scar suggests that scars were large and/or decayed, and/or that trees were growing slowly. Also, any tree with a detectable fire scar would have to be substantially older than 62 years.
Mortality was greatest among trees with complete crown scorch. Among trees with 100% crown scorch (which is different than complete topkill as defined by the author) mortality was greater for Q. wislizeni (75%) than Q. douglasii (17%). Data presented on the relationship between stem diameter and mortality show that for both species, almost all killed trees had diameters of 30 cm or less.
Post-fire sprouting. Epicormic sprouting in the canopy was more common among Q. douglasii (65% of all trees) than in Q. wislizeni (35%). Epicormic sprouting for both species was generally greater among trees with >50% crown scorching, except that in Q. wislizeni, trees with >95% crown scorch showed little epicormic sprouting. Most of the heavily scorched Q. wislizeni trees that survived apparently did so by producing basal sprouts. Two years after the fire, recovery of canopy cover among surviving trees was much lower for Q. wislizeni than for Q. douglasii.
Among trees (DBH>5 cm) which were completely topkilled (i.e., bole killed), survival of basal sprouts after two years was more common for Q. wislizeni (4/7 trees) than for Q. douglasii (2/7 trees). All of the 14 oaks in the smallest size class (>1.37 m height, < 5 cm DBH) in the fixed-area plots were topkilled and resprouted, but one subsequently died. Sprout height among these oaks was 10-50 cm 18 months after the fire, much smaller than their pre-fire heights.
Fire scarring. New scars developed on 59% of Q. douglasii, but many of these were not visible until 2 years after the fire. By contrast only 38% of the Q. wislizeni developed fire scars. What the author described as "old fire scars" were more prevalent on Q. wislizeni (86% of trees) than on Q. douglasii (47% of trees). It seems likely that all main stem and trunk cankers were described as old fire scars even though some (and perhaps many) of these could have been due to canker rot or other decay organisms. This could help account for the apparent discrepancy between new and old fire scars.
Other species. The author noted many T. diversilobum and some C. betuloides sprouts but no new shrub seedlings after the fire. Oak mistletoe (Phoradendron villosum) infections were reportedly killed when scorched.
This short paper presents a small portion of the data from the study reported in Haggerty 1994. It also includes overall tree mortality data for 11 other recently burned areas and one control (nonburned) area, compiled from an undisclosed number of 0.05 ha circular plots. All study areas were in the Kaweah River watershed and fire years ranged from 1979 to 1988. The dates of this survey are not stated, but they were presumably in 1988 and/or 1989. Total tree mortality, calculated from the live/dead tree density data presented in Table 3, ranges from 2% to 57% among the sites studied; mortality at the control site was 7%. Overall mortality for all trees was 30%, but most of this was apparently among species other than Q. douglasii. Mortality of Q. douglasii varied between 0.5% and almost 3% among 5 cm diameter size class groups. Slightly more mortality occurred in the smallest size classes (<10 cm DBH) and largest size classes (>40 cm DBH) compared with the intermediate size classes (10-40 cm DBH). Successful resprouting of Q. douglasii after topkill was almost entirely limited to trees with DBH of 25 cm or less.
Haggerty 1991a and 1994 are derived from data presented in this thesis. We reviewed the thesis some years ago and notes made at that time are the basis for the following comments.
Seedling densities. Seedling (<135 cm tall) density in the 13 locations described in Haggerty 1991a ranged from 0 to 130 /ha. Only one site had more than 32 seedlings/ha and 6 sites had less than 8 seedlings/ha. Seedling distribution was clumped, and did not appear to be correlated with site elevation or overstory density.
Fire and regeneration. The Ash Mountain fire did not have any discernable positive effect on blue oak seedling establishment or recruitment. At all sites, virtually all juvenile and small trees (DBH 0.1-5 cm) either predated the fires or arose as resprouts from small-diameter topkilled trees. At most sites, vegetative reproduction (resprouting) did not fully compensate for fire-related mortality.