Patterns of Distribution and Abundance of Corticolous Lichens and Their Invertebrate Associates on Quercus rubra in Maine

The distribution of corticolous lichens and their invertebrate associates was examined on the trunks of Quercus rubra L. at three sites in a mixed forest ecosystem in central Maine, USA. Thirty-two lichen species were found. Lichen biomass was greater on southern exposures, but species diversity was greater on northern exposures. Lichen assemblages contained terrestrial (Arthropoda) and aquatic (Rotifera, Nematoda, and Tardigrada) faunas. The aquatic fauna comprised 75% of the invertebrates retrieved from vacuumed assemblages. Oribatid mites comprised 86% of the terrestrial associates and macrolichen-dominant assemblages supported larger numbers of arthropods than crustose-dominant assemblages. The significant correlation between lichen biomass and Arthropoda, Tardigrada, and Rotifera abundance suggests that corticolous lichens are resources for invertebrate associates from these phyla. Corticolous lichens and invertebrates are important components of forest ecosystem function (Hale 1983; Hodgman & Bowyer 1985; Mattson 1977; Wallwork 1983). Biologists have long noted the association of invertebrates with corticolous lichens (Seaward 1988; Smith 1921). Invertebrate associates, which are invertebrates present in, under, on, or near lichens, belong to two faunas: Terrestrial and aquatic, with the aquatic fauna requiring moisture to be metabolically active. Terrestrial (Andre' 1985; Nicolai 1986) and aquatic (Meininger et al. 1985) faunas occur in corticolous lichens. Although several review articles (Gerson 1973; Gerson & Seaward 1977; Seyd & Seaward 1984) summarize numerous examples of lichen-invertebrate interactions, how and to what extent corticolous lichens and their invertebrate associates influence the health and productivity of forests is largely unknown. The primary objective of the present research was to determine whether a significant relationship exists between invertebrate associate abundance and the availability of corticolous lichens. Resource availability can be an important factor in determining organism abundance (Begon et al. 1986; Hart 198 1; Village 1987), and there are many ways that lichens can serve as resources for invertebrates, including food, camouflage, and oviposition sites. If corticolous lichen assemblages--characterized as groups of populations of co-occurring lichen species--function as important resources for invertebrates, then the abundance of invertebrates on tree trunks should correlate with lichen abundance. Another objective was to examine the distribution patterns of corticolous lichens and their invertebrate associates. Exposure can be an important factor for epiphytic distribution (Barkman 1958) and Quercus spp. are often rich in invertebrates (Morris 1974; Opler 1974; Southwood et al. 1982). Therefore, the distribution and abundance patterns of corticolous lichens and their invertebrate associates were compared on north and south exposures of the most important lumber species of oak in North America, Quercus rubra L. METHODS AND MATERIALS Study area and sites. -The University of Maine Dwight D. Demeritt Forest is in Penobscot Co., Maine (44?52'N, 68042'W). This area has a total annual precipitation of 99 0007-2745/89/453-460$0.95/0 This content downloaded from 207.46.13.18 on Fri, 13 May 2016 07:36:24 UTC All use subject to http://about.jstor.org/terms 454 THE BRYOLOGIST [VOL. 92 cm, which includes an annual snowfall of 1.93 m. The mean temperature for January is 7.90C and for July 19.90C (Baron et al. 1980). The Forest comprises four tracts of land totaling 1,746 acres. Approximately 1,599 acres are forested with softwood, mixed, and hardwood stands. Study sites were three uneven-aged mixed stands containing both Quercus rubra and Pinus strobus L. These sites are all within a 3 mile radius and were chosen in order to minimize intersite differences in abundance and distribution patterns due to dispersal and microclimate. These sites were designated Farm, Pens, and Cove. The Farm site, once part of a farm-homestead, has Dixmont stony and ledgy silt loam soils, which are shallow, well drained soils from glacial till (USDA Soil Conservation Service 1963). Dominant trees are Q. rubra, P. strobus, Acer saccharum Marsh., Fagus grandifolia Ehrh., and Tsuga canadensis (L.) Carr. Crown closure is approximately 70-80%. The Pens site soils, Buxton and Buxton-Biddeford, are deep, imperfectly drained clay-loam. Dominant trees are P. strobus, Q. rubra, and Acer rubrum L. Crown closure is over 90%, which makes the Pens the least illuminated of the sites. A jogging trail and a dirt service road intersect this site and a parking lot skirts the southwest sector. The Cove site is abutted by the Stillwater River. In contrast to the other sites, part of the Cove site is under water during spring snow melt. The soils belong to the Lovewell series, which are deep, moderately well drained soils of flood plains. Quercus rubra and A. rubrum are dominant throughout the site. Pinus strobus is codominant at northern and southern extremes of the Cove. Crown closure is approximately 80%. Lichen sampling. -A stratified random design was used. Sites were subdivided into topographic strata based on slope (Greig-Smith 1983). For each stratum at each site study trees were selected by matching the last two digits of calculator generated random numbers to sequentially numbered Q. rubra trees that had a diameter at breast height (dbh) of 16 cm or greater. This minimum dbh was selected to clarify the influence of aspect and to minimize the influence of tree size and age on diversity. Earlier researchers found 10 cm to be the critical dbh; the frequency of lichen species did not vary significantly as long as the dbh was 10 cm or greater (Culberson 1955; Hale 1955). Trees near potential sources of human disturbance, such as roads, were avoided whenever possible. Study trees were marked with aluminum tags. A compass was used to divide each study tree into four quadrats, one for each aspect. Quadrats varied in width, depending on the diameter of the tree, but had a constant height of 150 cm, which prevented vertical stratification effects (Hale 1952; Harris 1971). Lichen species found 30180 cm above the ground were recorded for each aspect. The three most dominant species, as determined by visual estimates of area covered, were recorded for north and south aspects for each tree. Invertebrate sampling.--Invertebrates were sampled in the previously identified lichen assemblages at the three sites in the summer of 1986. Terrestrial invertebrate sampling dates were 18, 19, 20 June and 23, 26, 27 August. Aquatic fauna were sampled 17 July and 23, 26, 27 August. Due to variability in size and location of lichen thalli, and observations of invertebrate motility within assemblages, I developed a consistent method of sampling invertebrates based on sampling the vegetation (lichens) within a fixed unit area for a fixed unit of time (Southwood 1978). Quadrats sampled for invertebrate associates were subsets of the initial north and south lichen assemblage quadrats so that invertebrate distribution patterns could be related to lichen assemblage composition. Thus, quadrats again extended from 30 cm to 180 cm above the ground, but were long, narrow strips (150 cm x 10 cm) centered at zero degrees for north and south aspects on each tree. Samples were gathered by vacuuming each 1,500 cm2 quadrat for 2 min. with a portable, commercial handvacuum equipped with the accompanying crevice attachment, which functioned as a scraper. The rigidity of the crevice attachment coupled with the suction of the vacuum made removal and retrieval of even crustose species of lichens possible without damaging the trees. Vacuumed samples were collected from northern and southern quadrats of each tree. All sampling days were sunny and all sampling was conducted from 9 AM to 1 PM. If a sample contained bark or nonlichen vegetation, it and the corresponding opposite aspect sample were discarded. Samples were stored in plastic bags and weighed later that same day. Air-dried sample weights were recorded and fauna extracted the same day they were collected. Invertebrate extraction and processing. -Terrestrial invertebrates were extracted from vacuumed lichen assemblage samples with improvised Berlese funnels (Borror et al. 1981). The relatively small amounts of lichen in individual samples, their already dry nature (they were collected dry), and concern over predation losses affecting sample composition necessitated a brief extraction period. Therefore, a 75 W tungsten light bulb was illuminated 20 cm above individual samples for 5 min. Terrestrial invertebrates were preserved in 70% ethyl alcohol. Lichen samples were retained in plastic bags. To enhance accuracy, terrestrial invertebrate abundance was determined by counting fauna in both alcohol and the retained dried lichens for each sample. Aquatic invertebrates were extracted by soaking lichen samples in 100 ml distilled water in sterile glass containers for 4 hr. Earlier extraction trials with longer incubations indicated high tardigrade and rotifer mortality and carnivory losses. Aquatic invertebrates were retrieved by filtering soaked samples through a 250 Aim mesh screen and then rinsing residual lichens with 75 ml distilled water to enhance microinvertebrate retrieval prior to preservation with 10% formalin. Seventy-eight paired samples from north and south aspects of 39 trees were examined: 46 for terrestrial invertebrates and 32 for aquatic invertebrates. Invertebrates were sorted to higher taxonomic groups and counted under a dissecting microscope. Data analysis. -Data analyses included the nonparametric Wilcoxon Paired Sample Test and the Mann-Whitney Test (Zar 1984). The significance level was 0.05 for all analyses.