The Nature Conservancy, Boulder, Colorado, United States of America.
Department of Geography, University of Colorado Boulder, Boulder, Colorado, United States of America.
PLoS One. 2020 Jan 15;15(1):e0226926. doi: 10.1371/journal.pone.0226926. eCollection 2020.
Forested fire refugia (trees that survive fires) are important disturbance legacies that provide seed sources for post-fire regeneration. Conifer regeneration has been limited following some recent western fires, particularly in ponderosa pine (Pinus ponderosa) forests. However, the extent, characteristics, and predictability of ponderosa pine fire refugia are largely unknown. Within 23 fires in ponderosa pine-dominated forests of the Colorado Front Range (1996-2013), we evaluated the spatial characteristics and predictability of refugia: first using Monitoring Trends in Burn Severity (MTBS) burn severity metrics, then using landscape variables (topography, weather, anthropogenic factors, and pre-fire forest cover). Using 1-m resolution aerial imagery, we created a binary variable of post-fire conifer presence ('Conifer Refugia') and absence ('Conifer Absence') within 30-m grid cells. We found that maximum patch size of Conifer Absence was positively correlated with fire size, and 38% of the burned area was ≥ 50m from a conifer seed source, revealing a management challenge as fire sizes increase with warming further limiting conifer recovery. In predicting Conifer Refugia with two MTBS-produced databases, thematic burn severity classes (TBSC) and continuous Relative differenced Normalized Burn Ratio (RdNBR) values, Conifer Absence was high in previously forested areas of Low and Moderate burn severity classes in TBSC. RdNBR more accurately identified post-fire conifer survivorship. In predicting Conifer Refugia with landscape variables, Conifer Refugia were less likely during burn days with high maximum temperatures: while Conifer Refugia were more likely on moister soils and closer to higher order streams, homes, and roads; and on less rugged, valley topography. Importantly, pre-fire forest canopy cover was not strongly associated with Conifer Refugia. This study further informs forest management by mapping post-fire patches lacking conifer seed sources, validating the use of RdNBR for fire refugia, and detecting abiotic and topographic variables that may promote conifer refugia.
森林火灾避难所（能够在火灾中幸存的树木）是重要的干扰遗产，为火灾后再生提供种子来源。最近的一些西部火灾，特别是在黄松（Pinus ponderosa）森林中，限制了针叶树的再生。然而，黄松火灾避难所的范围、特征和可预测性在很大程度上是未知的。在科罗拉多前缘（1996-2013 年）的黄松主导的森林中，我们评估了避难所的空间特征和可预测性：首先使用监测燃烧严重程度趋势（MTBS）燃烧严重程度指标，然后使用景观变量（地形、天气、人为因素和火灾前森林覆盖）。使用 1 米分辨率的航空图像，我们在 30 米网格单元内创建了一个二进制变量，即火灾后针叶树存在（“针叶树避难所”）和不存在（“针叶树不存在”）。我们发现，针叶树不存在的最大斑块大小与火灾大小呈正相关，并且 38%的燃烧面积距离针叶树种子源≥50m，这表明随着气候变暖，火灾规模进一步增大，限制了针叶树的恢复，这是一个管理上的挑战。在使用两个 MTBS 生成的数据库（专题燃烧严重程度类（TBSC）和连续相对差分归一化燃烧比（RdNBR）值）预测针叶树避难所时，TBSC 中低和中等级别的先前森林区域的针叶树不存在的比例较高。RdNBR 更准确地识别了火灾后的针叶树存活率。在使用景观变量预测针叶树避难所时，在最高温度较高的燃烧日，针叶树避难所的可能性较小：而在更潮湿的土壤上，在靠近较高等级溪流、房屋和道路的地方，以及在较不崎岖的山谷地形上，针叶树避难所的可能性更大。重要的是，火灾前森林冠层覆盖与针叶树避难所没有很强的相关性。本研究通过绘制缺乏针叶树种子来源的火灾后斑块，进一步为森林管理提供了信息，验证了 RdNBR 在火灾避难所中的使用，并检测了可能促进针叶树避难所的非生物和地形变量。
