Historical spatial patterns and contemporary tree mortality in dry mixed-conifer forests

Management and restoration of the dry, frequent-fire forests of the North American west depend on sound information about both historical and contemporary conditions to adequately address repercussions of fire suppression and changing climate. The purpose of this study is to quantify historical tree spatial patterns and assess recent mortality trends for old and large tree populations in dry mixed-conifer forests of the Northern Rocky Mountains. We analyzed historical reconstructions of forest spatial structure across six 1.0ha plots located in mixed ponderosa pine/Douglas-fir stands in western Montana, USA. Across plots, 10-23% of trees occurred as widely spaced individuals (no neighbors within 6m), with the remaining 77-90% of trees occurring in clumps (groups of two or more trees spaced less than 6m apart). Mean clump size was 2.2-4.2 trees per clump, although large clumps (>10 trees) were common. Global spatial analysis with the pair correlation function indicated that ponderosa pine patterns were spatially random at all scales, while Douglas-fir trees were spatially aggregated at scales less than 6m. The proportion of plot area farther than 9m from the nearest tree ranged from 1% to 20% across the six study plots. Mortality rates between 1991 and 2012 averaged 0.8%yr^-1 for old ponderosa pine and 2.1%yr^-1 for old Douglas-fir. We found limited evidence of density-dependent mortality for both species pooled and for ponderosa pine individually. Douglas-fir that died between 1991 and 2012 had higher local Stand Density Index (SDI) of Douglas-fir neighbors than did Douglas-fir that survived (P=0.003), indicating conspecific density-dependent mortality. When compared to ponderosa pine and dry mixed-conifer forests in other regions, trees were distributed much more evenly across clump sizes in our Montana study sites. Our analysis provides an estimate of the historical range of variability for spatial aspects of forest structure in dry mixed-conifer forests of the northern US Rockies and is relevant to the design of restoration and climate change adaptation treatments in such forests.