Taxonomic calibration uses environmental variables, such as mean annual temperature and mean annual precipitation, to calibrate mathematically the relationship between climate and the taxonomic composition of modern vegetation. The taxonomic composition of extant taxa found within the fossil assemblage can then be compared with the modern taxon-climate calibration to infer paleoclimate. Taxonomic calibration is usually performed at the species level. Boyle, Meyer, and Enquist (2008) used taxonomic calibration at the genus and family levels to identify the closest modern analog for Florissant fossil flora and to infer paleoclimate. The study used the taxonomic composition of 241 modern forest plots to calibrate multiple climate variables with forest type. At both the genus and family level their analysis placed the Florissant fossil flora between modern deciduous forests of the eastern United States and the humid subtropical montane forests of central and northeastern Mexico. Estimates for the mean annual temperature for Florissant during the Eocene using their calibration with modern forest plots ranged from 14.7 ± 2.2 degrees Celsius using the genus level to 15.6 ± 2.5 degrees Celsius using the family level. This range of temperatures is consistent with a warm temperate lowland or subtropical to tropical highland climate (Boyle et al, 2008, p. 42). This study demonstrates the possibility of using taxa higher than the species level as paleoecological and paleoclimatic indicators. NLR and taxonomic calibration at the species level require precise identification of taxa. Taxonomic calibration using higher taxa requires only approximate identifications to the family level.

Leaf Morphology

Leaf physiognomy refers to estimating paleoclimate using leaf morphology. Leaf physiognomy has the advantage of not requiring plant identification and is said to be ataxonomic. Leaf physiognomy analysis takes advantage of the fact that extant angiosperm dicots exhibit certain leaf structures, which correlate to precipitation, humidity, and temperature. Convergent evolution produces similar leaf adaptations for similar environmental conditions among flowering plants of different lineages. For example, evergreen leaves in humid environments usually have drip-tips, compound leaves are frequently associated with deciduous forests, and leaves with serrated margins dominate humid, cool environments while leaves with entire margins prevail in humid, warm environments (Stewart & Rothwell, 1993, p. 494). Eight leaf characters often used in this approach include leaf size distribution, leaf margin type, drip tips, organization (simple or compound), venation-pattern, venation density, leaf texture, and leaf base type (Cleal & Thomas, 2009, p. 34). Two methods utilize leaf physiognomy to reconstruct paleoclimate: Climate-Leaf Analysis Multivariate Program (CLAMP) and Leaf Margin Analysis (LMA).

Climate-Leaf Analysis Multivariate Program

Climate-Leaf Analysis Multivariate Program (CLAMP) utilizes a data base that correlates modern vegetative types to climate for estimating paleoclimatic variables among fossil leaf assemblages. CLAMP uses 31 leaf character states of at least 20 species of woody dicots to map out the vegetation within a small area associated with a climate station (Wolfe, 1995, p. 122). The modern data base now represents 173 plant communities mostly from Northern American forests. Leaves of a fossil assemblage can be scored using the same 31 leaf character traits and positioned on the physiognomic space defined by present day plant communities. In this way, CLAMP can provide climatic parameters related to precipitation, humidity, and temperature. Scoring the multiple characteristics used for CLAMP requires expertise and may be challenging for fossil leaves.

Leaf Margin Analysis

Leaf Margin Analysis (LMA) is a univariate approach which correlates leaf margin (entire vs. toothed) with mean annual temperature (MAT). Warmer climates have a higher percentage of smooth-edged species than cooler climates. Another univariate approach correlates leaf surface area with mean annual precipitation (MAP). Leaves tend to be small in hot, dry climates and larger in wetter climates. A univariate approach only allows you to evaluate one parameter of climate at a time. However, scoring one character trait at a time may make a univariate approach less ambiguous (Wilf, 1997, p. 385).

Botanical Estimates

Modern Florissant has a MAT of around 4 degrees Celsius and an annual precipitation of 38 centimeters (Meyer, 2003, pp. 51 & 52). How does the present day MAT and annual precipitation compare with Florissant during the latest Eocene as estimated by paleobotanical methods? NLR, taxonomic calibration, and CLAMP have been used to estimate the paleoclimate of Florissant. The CLAMP estimates of MAT range between 10.7 and 12.8 degrees Celsius (Meyer, 2001, p. 210). In general, this range is consistent with a temperate climate and, at the low end, implies a high frequency of freezing temperatures during the winter months. Estimates based on NLR give a MAT of around 18 degrees Celsius. In general, this mean annual temperature is consistent with a warm temperate climate that boarders on subtropical. MacGinite used the composition of the fossil flora to estimate a MAT of 65 degrees Fahrenheit or 18.3 degrees Celsius (MacGinite, 1953, p. 57). The addition of a palm leaf fossil to the Florissant fossil flora and an analysis of fossil pollen and spores add support for a warm temperate climate that was relatively frost free (Leopold & Clay-Poole, 2001, p. 29). The MAT estimates based upon taxonomic calibration fall between the cool temperatures predicted by leaf physiognomy and the higher temperatures predicted by the nearest living relative method (Boyle et al., 2008, p. 33). Smaller leaves with serrated margins are associated with seasonally dry climates. Fossil leaves at Florissant have been used to estimate an annual precipitation of 50-80 cm per year with a dry season (Leopold & Clay-Poole, 2001, p. 48; Meyer, 2003, p. 52).

Using Extant Insect Taxa to Infer Past Climate

Like modern plants, the distribution of many insect taxa correlates well with climate. Three methods have been developed that utilize the ecological tolerances of extant insect taxa to infer past climate and include; the modern analog approach, the nearest living relative approach (NLR), and the mutual climatic range (MCR) approach. All three methods have been used primarily for estimating the paleoclimate of Quaternary fossil assemblages.

Modern Analog & Nearest Living Relative

The modern analog approach uses the composition of extant climate indicator insect species found within a fossil assemblage to infer the paleoclimate of the ancient community. The NLR approach uses the climatic tolerances of modern insects believed to be the nearest living relatives of extinct taxa within the fossil assemblage to infer paleoclimate. Modern analog and NLR approaches are qualitative in that they have no specified number or method for selecting taxa to be used in a study. The MCR approach is quantitative and typically uses the overlapping climatic ranges of all extant, non-phytophagous beetle species within a fossil assemblage to infer paleoclimate.

Mutual Climatic Range Estimates