Examples from Two Modern Grid-Point Assemblages

To illustrate the MCR approaches and structure of the datasets, example results are presented for two modern grid-point assemblages (xeric and Taiga) used in Thompson and others (2008, 2012) (see locations in fig. 1). In these previous studies, we used four sample assemblages, but two of these assemblages included too many taxa to use as visual displays of the approaches.

The first example is from the xeric ecoregion (Ricketts and others, 1999) located in the Sonoran Desert of southern Arizona. There are ten species present at this grid point, nine of which are restricted to the warm deserts along the border with Mexico: Acacia gregii (catclaw acacia), Carneigia gigantea (saguaro), Cercidium floridum (blue palo verde), C. microphyllum (foothill palo verde), Chilopsis linearis (desert willow), Larrea divaricata (creosote bush), Olneya tesota (desert willow), Opuntia fulgida (jumping cholla), and Yucca elata (soaptree yucca). The tenth species, Salix exigua (coyote willow), is at the southernmost tip of a large range that extends northward to Alaska and eastward as far as the State of New York (Little, 1976, 1977).

For the MCR_un analysis, the rows in our tables represent the midpoints of 0.1°C increments of MTCO. Table 2 is a simplified version of the table for the xeric assemblage where, for brevity, only every fifth value is shown. The species present in the xeric grid-point assemblage are shown in the columns with a value of one indicating that the species occurs today in the temperature range represented by the row under consideration and zero indicating that it does not occur in that temperature range. The last column on the right is the sum of the ones and zeros for all species for the temperature range represented by a row. In this case, 10 is the maximum value because there are 10 species present in the assemblage, and the yellow area of the table indicates the temperature range where all 10 species co-occur (MTCO of 6.5°C to 12.5°C). The midpoint of this range is 9.4°C, which is very close to the observed value of 9.7°C.

A simplified version of the data table for the MCR_wt analysis of MTCO for the xeric site is also presented (table 3; again showing only every fifth row). This is the same matrix as in table 2, except that instead of only denoting presence with a one and absence with a zero, the cells contain the weights for a given taxon following the translation scheme described in the previous section. The far right column in table 3 lists the sums of weighted values; the yellow area indicates the MTCO range of complete overlap from table 2, and the gray area indicates the two rows with the highest sums of weights (MTCO of 10°C to 10.5°C) with a resulting estimate of 10.3°C.

The MCR_un and MCR_wt approaches for the estimation of MTCO at the xeric site can be shown graphically (fig. 6). In the left panel of figure 6, red histograms illustrate the weighted values for the ranges of MTCO where each species in the xeric assemblage occurs today, and the silhouette in the upper box shows the sums of these weights. Among the taxa in the assemblage, only Salix exigua can survive today under cold winter temperatures, and correspondingly the sum of weights is very low where MTCO is less than 0°C. The silhouette rises between MTCO of 0°C and ˜10°C as more species are represented where winter temperatures are less limiting; above ∼10°C, the silhouette declines and then drops precipitously above 15°C as winter temperatures rise above the habitable range for the species in the assemblage. The vertical yellow band represents the range of MTCO where all species co-occur (the MCR_un range), and the narrower gray band illustrates the range with the highest sums of weights. The observed (Obs) value and MCR_wt estimate are marked at the top of the diagram.

In the xeric example, both the MCR_un and MCR_wt approaches result in estimates that are very close to the observed value, but this is not true for all assemblages on the North American grid. The right panel on figure 6 illustrates such an assemblage, with a floristic list of eight species from a grid point in the Taiga ecoregion of northern Quebec (Ricketts and others, 1999). The vegetation here is typical of the northern boreal forest and includes Betula papyrifera (paper birch), Juniperus communis (common juniper), J. horizontalis (creeping juniper), Larix laricina (tamarack larch), Picea glauca (white spruce), P. mariana (black spruce), and Sorbus decora (northern mountain ash). These species live together today in a broad latitudinal band from northern Canada south to the northern reaches of Alaska, as shown in the blue histograms on figure 6, and are adapted to very cold winter temperatures. The MCR_un and MCR_wt MTCO ranges are both much wider than for the xeric assemblage, and the observed value of –24.4°C falls at the lower MTCO limit of the MCR_un range. Consequently, the MTCO estimates obtained from the MCR_wt and MCR_un approaches (–19.1°C and –14.2°C respectively) are significantly warmer than the observed value.

The comparison of the xeric and Taiga examples in figure 6 provides a lesson for those attempting to estimate climatic parameters from biologic assemblages. At the xeric site, MTCO has a strong controlling influence on the distributions of most of the species in the assemblage, and as a result, the MCR approaches result in estimates that are very close to the observed value. In contrast, the plant species in the Taiga assemblage are adapted to cold winter temperatures, and MTCO has much less influence on their distributions; the MCR approaches result in very broad MTCO ranges for the complete overlap of these species.