Reconstructions of fire rotation (FR) for high-severity fire in historical dry forests of the western USA, with corroborating evidence from sedimentary charcoal studies.

Studies are arranged by length of the fire rotation. Estimates from GLO data, FIA data, and early aerial photographs are shown in bold italics to emphasize their higher precision, while corroborative, less certain estimates from charcoal records are shown in regular type. The range of estimates in bold is used as the reference in this study.1 Methods for reconstruction included using charcoal data from sediment, using early aerial photographs or historical records, using the GLO tree data and a calibrated model [28]. I did not use the GLO line data’s direct records of entry and exit in burned areas, as these records represent moderate- to high-severity fires, not exclusively high-severity fires [40].2 These are intervals between severe fire episodes evident in alluvial deposits, that could approximate high-severity fire rotations, but are uncertain since area burned is not known and fire severity is more approximately reconstructed than with other methods. I considered data for the last 500 years from each paleo-environmental study, but also included in parentheses the interval between episodes in the last 2000 years, where this is available.3 These authors indicate that it is difficult to determine fire severity from their methods, and only identify the recent fire frequency as 3 per 1000 years, but they indicate that the documented fire episodes were followed by up to 100 years of recovery, which does suggest severe fires, although this is my interpretation.4 Fitch [32] suggested that low-severity fire dominated from 870 cal yr BP, but explains the possibility, but uncertainty, of a severe fire around 400 cal yr BP (p. 40), thus I include this single event, with a question mark, for the 500-year estimate. More certain is evidence of 3–5 severe fires in the last 2000 years (p. 42), but those all preceded 870 cal yr BP.5 These authors were not focused on counting the number of fire-episodes over the last 2000 years, thus I roughly estimated this from Fig 5B in Pierce and Meyer [33] as between 7–13, as there are 7 broad peaks in this figure, but they also report 9 major debris flows between about 950 and 1150 AD, thus the total could reach as many as 13. No severe fires occurred in the last 500 years.6 This author provided data on the number of watersheds, out of six sampled, that burned in high-severity events [36]. I used these data to approximate a high-severity fire rotation using the standard formula: period of observation / fraction of area burned. Thus, for the last 550 years, a total of 7 watersheds burned, thus the fire rotation is 550 / (7/6) = 471 years. And for the last 2000 years, a total of 18 watersheds burned, thus a fire rotation of 667 years. However, Bigio indicates that sample locations may be high in a watershed, thus it is not known that the whole watershed burned. This leaves these estimates as minima, which I have indicated by using “>” before the estimate.7 These authors identify periods of severe-fire activity after c. 1800 cal, yr BP, a peak in 800–500 cal yr BP, and at least one large, severe fire in the last 400 years, thus perhaps 3 episodes in the last 2000 years and one in the last 500 years. However, this is my approximation from their data, as they do not report recurrence intervals for severe fire.Reconstructions of fire rotation (FR) for high-severity fire in historical dry forests of the western USA, with corroborating evidence from sedimentary charcoal studies.

本研究纳入的各项案例按火灾轮回期（fire rotation）的长度排序。其中，源自通用土地管理局（General Land Office, GLO）数据、美国森林资源清查与分析（Forest Inventory and Analysis, FIA）数据以及早期航空摄影的估算结果以粗斜体标注，以突出其更高的精度；而来自木炭记录的佐证性、确定性较弱的估算结果则以常规字体呈现。本研究以粗体标注的估算区间作为参考基准[1]。本次重建采用的方法包括：利用沉积物中的木炭数据、早期航空摄影或历史记录，以及GLO林木数据与校准模型[28]。本研究未使用GLO线状数据中关于火烧区域进入与退出的直接记录，因该类记录仅表征中至高烈度火灾，而非仅高烈度火灾[40]。2 此类间隔指冲积沉积物中显现的高烈度火灾事件之间的时间跨度，可近似表征高烈度火灾轮回期，但由于火烧面积未知，且火灾烈度的重建精度相较于其他方法更低，因此该估算存在不确定性。本研究采用了每项古环境研究中近500年的相关数据，同时在括号内补充了可获取的近2000年内的火灾事件间隔数据。3 上述研究指出，其方法难以直接判定火灾烈度，仅将近期火灾频率估算为每1000年3次；但研究记录的火灾事件后通常伴随长达100年的恢复期，这暗示该火灾为高烈度火灾——该结论属于本研究的解读。4 菲奇（Fitch）[32]提出，公元870校准年BP（cal yr BP）以来以低烈度火灾为主，但同时指出约公元400校准年BP左右存在高烈度火灾的可能性（存在不确定性）[p.40]，因此本研究将该单次事件以带问号的形式纳入近500年的估算中。更为确定的证据显示，近2000年内共发生3~5次高烈度火灾[p.42]，但此类火灾均发生在公元870校准年BP之前。5 上述研究并未聚焦于统计近2000年内的火灾事件次数，因此本研究根据皮尔斯与迈耶（Pierce and Meyer）[33]的图5B进行了粗略估算：该图中共出现7个宽峰，因此火灾事件次数介于7~13次之间；同时该研究报告了公元950年至1150年间共发生9次大型泥石流，因此总次数最高可达13次。此外，近500年内未发生高烈度火灾。6 该作者提供了6个采样流域中发生高烈度火烧事件的流域数量数据[36]。本研究采用标准公式近似估算高烈度火灾轮回期：观测时长 / 火烧面积占比。据此，近550年内共有7个流域发生火烧，因此火灾轮回期为550 / (7/6) = 471年；近2000年内共有18个流域发生火烧，因此火灾轮回期为667年。不过比吉奥（Bigio）指出，采样点位可能位于流域的高海拔区域，因此无法确认整个流域均被火烧，因此该估算结果仅为最小值，本研究通过在估算值前添加">"符号予以标注。7 上述研究识别出公元约1800校准年BP之后的高烈度火灾活动期、公元800~500校准年BP的火灾活动峰值，以及近400年内至少1次大型高烈度火灾，因此近2000年内或存在3次火灾事件，近500年内或存在1次。不过，由于该研究未报告高烈度火灾的复发间隔，上述结论属于本研究基于其数据的近似推断。本研究重建了美国西部历史干旱森林的高烈度火灾轮回期（fire rotation, FR），并辅以沉积木炭研究的佐证性证据。