whether they represent the same leaf population, (2) assess whether a combined collection yields 24 more precise climate estimates, and (3) reconstruct paleoclimate to assess the character of 25 regional Early Eocene precipitation seasonality. Combined, the two samples yielded 43 dicot leaf 26 morphotypes. Analysis of leaf size distribution using ANOVA showed no difference between the 27 2 samples and they were combined for climate analysis. Climate analysis using leaf
47Early Eocene fossil floras from British Columbia and Washington State, coined the
48Okanagan Highlands, provide a window into the climate of the Early Eocene at mid-latitudes
49(e.g., Smith et al. 2010Smith et al. , 2012 Archibald et al. 2011 Archibald et al. , 2014, where modelling has identified 50 highly seasonal precipitation in the North American continental interior and ever-wet regimes on 51 the west coast and adjoining area of British Columbia (Huber and Goldner 2012 2005; Smith et al. 2010 Smith et al. , 2012. The Brandon University (BU) collection was made by S.B.
101Archibald in the early 2000s. Neither collection was made within a site stratigraphy (J.F.
102Basinger and S.B. Archibald, pers. comm., Sept. 4, 2015). The McAbee main site (Fig. 2) is an 1978; Greenwoood 1992 Greenwoood , 2007 Steart et al. 2002). These same patterns have been seen 141 in Eocene lake sediment leaf assemblages over short stratigraphic distances of <5 m (e.g.,
142Greenwood 1991; Smith et al. 2009, 2012). At Falkland, Smith (2011 Smith et al. 2009 Smith et al. , 2012 by changes in annual precipitation amounts over time (Smith et al. 2009(Smith et al. , 2012 or 'ever-wet' climate regime.
167
Materials and Methods
168
Fossil Flora
169The first of the two fossil samples examined in this study was collected by S.B.
170Archibald in the early 2000s and is currently curated at Brandon University (BU). This collection populations. The TRU collection was not available for analysis.
183
Leaf Morphotyping
184The fossil leaves of both the BU and US collections were separated into distinctive in depth taxonomic analysis of the fossil flora (Ellis et al. 2009;Yang et al. 2011 Yang et al. , 2015.
189Important morphological characteristics considered in this process include: leaf size, margin 190 type, apex shape, base shape, and venation. Each morphotype is defined using the character 191 scores of multiple leaves representative of the morphotype, entered into the MLA pro forma.
192Scores of leaf size class for each leaf morphotype were made using both the leaf size classes Yang et al. 2011Yang et al. , 2015. collections was directly measured using the equation:
203(1) Leaf Area = Length × Width × 0.67 (Cain and Castro 1957; Greenwood 1992).
204The leaf area data was analyzed using two one-way ANOVA tests in Microsoft®-Excel to generate an estimate for MAP. In addition, the smallest and largest specimens from each 229 morphotype were directly measured using equation (1) could not be directly measured were excluded from the analysis, although future work will ...