Sexual Fecundity is Correlated to Size in the Lichenized Fungus Xanthoparmelia cumberlandia

There are no standard measures of fungal fitness, and yet descriptions of natural selection in fungi require an understanding of how to compare the success of two individuals. Success, or fitness, is normally understood to be a combination of survival and reproduction. Xanthoparmelia cumberlandia is a sexual, lichenized fungus. By recording the size of, and number of sexual structures on, individual lichens we demonstrate a significant correlation between size and reproductive effort in this species, showing that size is an easily measured surrogate of fitness. Published data of other lichen species, for example Umbilicaria spodochroa or Xanthoria parietina, also show a correlation between size and sexual fecundity, indicating that the correlation may be a general feature of sexual lichens. However, patterns of resource allocation differ between lichen species. Published data collected from U. spodochroa are linear, demonstrating that larger lichens allocate equivalent resources to growth and reproduction. In contrast, the data of X. cumberlandia are curved, indicating that in this species larger lichens allocate a disproportionate share of resources to reproduction. Organisms are variable, and the variability causes one individual to be more successful than another; because the variation is heritable a lineage can evolve. To understand the evolution of fungi we first must understand how to measure differences in the success of two individuals. Success is normally understood as fitness, but there are no standard measures of fitness in fungi (Pringle & Taylor 2002). Fitness may be quantified as the combination of survival and reproduction (Day & Otto 2000). If we are to understand basic hypotheses related to the adapative significance of fungal phenotypes, for example why a species is sexual or asexual, we first must find useful ways to measure and understand which fungal individuals are more or less fit within a habitat. Lichens are symbiotic associations of fungi and either algae or cyanobacteria; individual lichens are easily counted or measured. The fitness of an individual, sexual lichen might be measured as female fecundity, by counting the number of female sexual structures (ascogonia), or as male fecundity, by counting the number of male sexual structures (pycnidia). But these structures are hardly visible and so counting ascogonia or pycnidia can be difficult or impossible. Sexual spore-bearing structures, the apothecia, are formed by the fusion of female and male sexual propagules. Apothecia are more easily counted than ascogonia or pycnidia, and although no data relate apothecial numbers to spore numbers for any species of lichen, it is reasonable to assume that there is some correlation between the two parameters. But counting apothecia can be time consuming, especially when an individual lichen possesses hundreds or even a thousand apothecia. In contrast, measuring the size of an individual lichen is straightforward, but useful only if apothecial number is correlated to size. In this study we demonstrate than an aspect of fungal fitness, apothecial number, is correlated to the size of a fungal individual. Xanthoparmelia cumberlandia is a sexual, lichenized fungus. Apothecia are retained on the thallus of the lichen for the lifetime of an individual; an apothecial count is a direct measure of the reproductive effort expended by a lichen in its lifetime. By counting apothecia and measuring the size of X. cumberlandia individuals, we demonstrate that this aspect of fungal fitness is correlated to the size of an individual. Sizes are easily measured and for this reason may serve as a useful estimate of fungal fitness. In fact the technique is analogous to measuring the biomass of a plant when biomass is a predictor of fruit and seed number—this is a methodology that is commonly employed by plant population biologists. MATERIALS AND METHODS Species. Xanthoparmelia cumberlandia is a foliose lichen commonly distributed throughout California and western (excluding Nevada) and eastern North America (Fig. 1). Apothecia are common, and appear as cupshaped structures at the center of the thallus. Apothecial discs are darkly pigmented, although the margins are typ222 [VOL. 106 THE BRYOLOGIST FIGURE 1. Xanthoparmelia cumberlandia growing in the University of California Botanical Garden. ically toothed, rolled, and the same color as the thallus. Pycnidia are also common, and appear as black dots on the surface of the thallus. Ascogonia are not visible to the unaided eye. The lichen X. cumberlandia does not create specialized asexual propagules e.g., isidia or soredia. The species is usually found on exposed or somewhat shaded siliceous and other noncalcareous outcrops or boulders in open and dry habitats. Descriptions of the species can be found in Brodo et al. (2001) and McCune and Geiser (1997). Study site and sampling. Lichens measured in this study were found within the University of California at Berkeley’s Botanical Gardens during June and July of 2002. A total of 101 lichens were sampled from the ‘‘Southern Africa’’, ‘‘New World Desert’’, and ‘‘California’’ displays on rocks that are densely populated with individuals of X. cumberlandia. We chose to measure lichens of a variety of sizes, and included only those individuals whose boundaries were easily defined. Measures of area. Transparent plastic sheets were taped over an individual lichen and the perimeter of the lichen traced using an ultra fine point permanent marker. Each of the lichens was traced by only one of the authors (D. Chen) so that a consistent technique was employed over the course of the study. Tracings were scanned and files opened in Adobe Photoshop 6.0.1 (Adobe Systems Inc. 2001). The interior of each tracing was painted black using the ‘‘paintbucket’’ command. Filled images were saved as ‘‘Tagged Image File’’ or ‘‘tif’’ files. Images were subsequently measured with NIH Image 1.62 (a program available from the NIH website, rsb.info. nih.gov/nih-image/download.html). Measures were taken by counting the number of black pixels in each image and converting the count to a standard measure of area, in this case, cm2. For details on using NIH Image 1.62 to gather morphometric data see Ruzin (2002; available from microscopy.berkeley.edu). After tracing each lichen, a spray-bottle of water was used to wet the lichen and open apothecia. A digital photo was taken and used to count apothecia. Apothecial counts. One of two methods was used to count apothecia. Initially, the number of apothecia on each X. cumberlandia was determined visually by taping writeon transparency film over the photo of an individual lichen. Apothecia were simultaneously marked with a fine point permanent marker and counted with a hand-tally counter. Marking each apothecium as it was counted ensured that a single apothecium was not counted twice. Very large lichens, and those that grew around the curves of rocks, were difficult to fit into a single picture. In these cases thread was used to halve or otherwise divide the lichen and each piece photographed and counted separately. Our technique was refined when we discovered that instead of using a permanent marker to mark apothecia we could open the photos in Adobe Photoshop version 6.0.1 (Adobe Systems Inc. 2001) and use the mouse and ‘‘airbrush’’ tool to mark each apothecium. By this technique, the number of apothecia was still recorded with a handtally counter. The advantage of using Adobe Photoshop to mark apothecia was the ability to enlarge photos as we counted so that apothecia growing closely together were more easily distinguished. Analyses. Data were first explored with a simple correlation statistic using JMP version 3 (SAS Institute Inc. 1999). We used maximum likelihood methods to determine the relationship between fecundity (the number of apothecia, Ai) and size (Si). We compared two models: a linear model (where the predicted fecundity Âi of individual i increases 2003] 223 PRINGLE ET AL.: SEXUAL FECUNDITY FIGURE 2. Fitting a curved model to data of individual X. cumberlandia. — A. Data. B. The model Âi 5 22.8 1 0.78(Si). Dashed lines are 95% confidence intervals. FIGURE 3. Individuals that measure less than 10 cm2 possess few, if any, apothecia. linearly with size Si): Âi 5 b 1 mSi and a non-linear model (in which larger individuals allocate proportionately more or less to fecundity): Âi 5 b 1 m(Si). Fitted parameters include an intercept (b), the slope of the relationship between fecundity and size (m) and exponent e. Values of e greater than one would indicate that larger individuals allocate proportionately more to reproduction, values less than one would indicate that fecundity saturates with size. As the number of apothecia is a positive integer value we used a Poisson likelihood to estimate the parameter vector p (b, m, and if the non-linear model, e). The negative loglikelihood is thus: j ˆ ˆ L(A z p) 5 2A 1 A 3 log(A ) 2 log[G(A 1 1)] O i i i i i51 where A is the j length vector of observed apothecial counts (Ai’s). We used a likelihood ratio test to determine which model best fit the data. A likelihood ratio test compares nested models by determining whether the extra parameter (in our case, exponent e) significantly increases the model fit. Significance is assessed by the difference in likelihoods of the two models, which is distributed as a chi-square distribution with degrees of freedom determined by the difference in parameters (in this case, one; Hilborn & Mangel 1997). Maximum likelihood analyses were completed using the R statistical package (available from www.r-project.org).