}`HPLJ2%}.MT 1.00" .MB 1.00" .oj off .pn 3 ŠÜÀ´ð Figure 1. Map of giant sequoia groves on western slope of Sierra Nevada, California. The sampled groves are shown with arrows. .pa Ð/3  Š.pn 7 Figure 2. Map of fire scar collections in Giant Forest, Sequoia National Park. Solid triangles are locations of sampled trees. Three letter codes correspond to site names listed in Tables 1 and 2. Unshaded area is approximate boundary of the mixed conifer forest with sequoias. The dark shaded areas within the unshaded area are approximate locations of wet meadows. .pa +  2  Š Figure 3. Map of fire scar collections in Mariposa Grove, Yosemite National Park. Figure 4 . Map of fire scar collections in Big Stump Grove, Sequoia National Park. .pa € 2  Š.pn 9 Figure 5. Map of fire scar collections in Atwell Grove, Sequoia National Park Forest. The convex hull for all sampled trees is shown as a line connecting the outermost trees. Figure 6. Map of fire scar collections in Mountain Home Grove, Mountain Home State Forest. The convex hull for all sampled trees is shown as a line connecting the outermost trees. .pa € 2  Š.pn 11 Figure 7. A fire-scarred snag - "Pluto's Chimney" - in the Mariposa Grove. This ancient snag (tree no. MPU8) had an earliest crossdated ring at A.D. 554, but this specimen was far from the pith and the tree was probably many centuries older. The base of the tree was entirely com posed of buttresses and deep fire scar cavities. We extracted partial sections from near ground level at many locations around the stem. .pa - 2  Š.mt .5" Figure 8. Sampling sequoia stump with "eyesocket" fire scar in Big Stump Grove. The partial section removed from the center of this eyesocket contained 60 different fire scar dates. The former ground level was approximately at knee height on the sawyer (top). Numerous overlap ping fire scars can be seen within the eyesocket below the cut (bottom). .pa °1 2  Š.mt 1.0" .pn 23 Figure 9. Display section at General Sherman Tree parking lot in the Giant Forest. This section was taken near ground level and fire scars were visible extending along the edges of the convo luted cavities all the way into the center. This tree had 127 fire dates and a pith at 257 B.C. .pa ð-p5  Š.pn 25 Figure 10. Detailed portion of Circle Meadow master fire chronology, A.D. 1100 to 1600. Horizontal lines represent sampled time spans of different trees, and symbols denote fire dates determined by different fire indicators. This period illustrates the remarkably detailed fire history information contained in giant sequoias. Note the high frequency of fires during the period from ca. A.D. 1100 to 1300, and the perceptible decline in frequency after ca. 1300. .pa ð- 2  Š.pr or=p .rm 6.5" Figure 11. Sampled time spans of fire-scarred sequoia trees in the five groves. Each horizontal line represents the time span of dated material from each tree. Ten to 14 partial sections with fire scars were sampled per tree. .pa à. 2  Š.pn 28 Figure 12. Fire scars and growth releases. Fire scars were clearly identified on sanded sur faces. Large rings were usually observed at the margins of the old wounds where the healing process accelerated growth, however, increased growth ("growth releases") were also usually visible in rings following fire scar dates at points far removed from the wound boundary. .pa à. 2  Š.oj on .pn 34 Figure  13.  Frequency distributions of fire intervals - all fires included. Fire  dates  recorded  by fires  scars and other types of indicators during the period 500 to 1900 were used in  computing these distributions. See Table 4 for summary of means and variations. The last category of  fire intervals (far right) includes intervals greater than 20 years. .pa à. 2  ŠFigure  14.  Frequency  distributions of fire intervals - more than one  tree  recording  fires.  Fire dates  recorded by fire scars and other types of indicators on more than one tree  within  groves during the period A. D. 500 to 1900 were used in computing these distributions. See Table 4 for summary of means and variations. The last category of fire intervals (far right) includes intervals greater than 30 years. .pa ð- 2  Š.pn 45 .pr or=l .rm 9.5" Figure  15.  Master  fire  scar chronology and percent trees scarred per year in Circle Meadow. Fire dates recorded  by  fire  scars  only  are shown  in  this  graph. Note the higher frequency during the period A.D. 1000 to 1300, but smaller apparent size of fires  as  reflected  in  the numerous short bars in the histogram (bottom) of percentage trees scarred per year. .pa Ð/ 2  Š.pr or=p .rm 6.5" Figure  16. Fires per century (frequency) versus mean percentage of trees recording fires  (size) per century for the period A. D. 500 to 1800. .pa À0 2  ŠFigure  17. Centennial fire frequencies in five groves. Numbers of fire dates recorded  per  cen tury  are  plotted  for  fires recorded by all types of fire indicators, and for  fire  scars  alone.  The frequencies  are  plotted on the first year of the century. For example the value at  A.D.  1300  is the fire frequency for the period 1300 to 1399. .pa à. 2  Š.pn 54 Figure 18. Comparison of fire frequency variations among five groves. Centennial fire  frequen cies  (all  fire  indicators)  are plotted together in the upper graph. The  lower  graph  shows  the frequency of fires in all groves combined in moving periods. The upper line is a 50-year moving period, and the lower line is a 20-year period. See methods for further explanation. .pa à. 2  Š.mt 0.5" .pn 56 Figure  19.  Simulated fire frequencies at reduced sample sizes during the period A.D.  1000  to 1599. Each reduced sample size was randomly selected (i.e., trees removed from the data  set) 50 times, and the mean fire frequencies and 95% confidence intervals were computed for the 50 simulations.  The  upper  curves  in each plot were the  mean  fire  frequencies  and  confidence intervals  based  on  all  fire dates and the lower curves were based on  fires  recorded  by  more than one tree. .pa Ð/ 2  Š.pn 59 Figure 20. Simulated and observed fire frequencies in Circle Meadow. The upper plot is  based on all fires and the lower plot is based on fires recorded by more than one tree. The curves  are the mean fire frequencies and 95% confidence intervals at reduced sample sizes (simulated) for the base period A.D. 1000 to 1599. The solid dots are the observed fire frequencies and  corre sponding  maximum sample sizes during each century from the B.C./A.D. boundary (0)  to  the present  century  (1900s).  Centuries above or below the mean curves  are  interpreted  to  have relatively  higher  or  lower  (respectively)  fire frequencies than  the  base  period  fire  frequency adjusted for sample size. .pa ð- p5  ŠFigure  21.  Trends  in fire frequency among five sequoia groves based  on  the  comparison  of simulated  fire  frequencies at reduced sample sizes for the base period A.D. 1000 to  1599  and observed centennial fire frequencies. Positive trends were identified for observed centennial fire frequencies  greater than the simulated mean fire frequency curves , while negative trends  were identified for observed fire frequencies falling below the mean curves (See Fig. 20). .pa À0p5  Š.pn 71 Figure  22.  Summary  of  fire scar seasonality in five  groves.  Percentages  of  identified  intra- annual  fire scar positions since A.D. 500 are shown. E = earlywood (no  further  sub-classifica tion  is possible), EE = first one third of the earlywood, ME = middle one third of earlywood,  LE = last one third of earlywood, L = latewood, D = dormant (occurring on the ring boundary). .pa °1p5  ŠFigure  23.  Temporal  changes in fire scar seasonality. The percentage of  each  type  of  intra- annual fire scar position was identified for century long periods. .pa 3p5  Š.pn 76 .mt 0.75" .mb 0.75" Figure  24.  A.D. 1297 growth release. Ring widths in eight radial growth series from  giant  se quoias  growing  in Mountain Home Grove. The vertical dashed line shows the date  of  a  1297 fire  scar on most specimens sampled in this grove. The DSQ specimens were collected  by  A. E. Douglass early in this century. A large growth release was observed following the 1297 fire in these and nearly all other specimens we have collected from this grove.