Southeastern Naturalist 1 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 22001144 SOUTHEASTERN NATURALIST Vol1.3 1(31,) :N1–o2. 51 Rhododendron Decline in the Great Smoky Mountains and Surrounding Areas: Intensive Site Study of Biotic and Abiotic Parameters Associated with the Decline Richard Baird1,*, Alicia Wood-Jones1, Jac Varco2, Clarence Watson3, William Starrett1, Glenn Taylor4, and Kristine Johnson4 Abstract - Rhododendron dieback was continuously observed with increasing frequency on Rhododendron maximum (Rosebay Rhododendron) during the last 20 years in the southern Appalachian Mountains. The dieback was especially evident following several years of drought from 2004 to 2008 recorded in Great Smoky Mountains National Park (GRSM). With the concern that a disease epidemic could occur, a holistic study evaluated site factors including tree health, number of clonal units, aspect, slope, depth to bedrock, and rhizosphere microbes. This study was conducted at two locations: Laurel Falls in GRSM and Albert Mountain in Nantahala National Forest (NNF). Yearly sampling for nematodes showed no differences in frequencies across or between years. A total of 11 species were identified from replicated healthy and dieback plots with no significant trends observed. Criconemella xenophus, Helicotylenchus sp., and Meloidogyne sp. were the species most commonly found. Belonolaimus sp. occurred at the NNF site at below 1% of the total nematode population identified, but this nematode species is considered damaging to crops and forest nursery seedlings even at low numbers. Fungal/Oomycota diversity and densities were determined from roots and rhzosphere soil samples using three identification methods. The results ranged from 110 species of fungi to 0 for Oomycota. Of 110 fungi isolated, one putative root pathogen was identified, and the saprophytic species Mycena silvae-nigrae (unknown Basidiomycota 1) was the most common match using the GenBank database. Elevation at NNF was significantly greater than at GRSM, with significantly greater dieback levels at the higher elevation. Furthermore, greater dieback ratings were associated with significantly greater tree diameters. No trends were observed for percent slope or nutrient levels when compared between healthy and dieback sites or locations. Site factors such as aspect, elevation, associated nematode species, and a putative root pathogen may form a disease complex resulting in Rosebay Rhododendron dieback. Introduction The Great Smoky Mountains include a wide range of temperate to boreal forest types (Brown 2000). In the more acidic sites, between 305 and 1676 m elevation, Tsuga canadensis L. (Eastern Hemlock) grows on the slopes, but is most commonly found with Liriodendron tulipifera L. (Tulip Poplar) in the coves that contain an 1Department of Biochemistry, Molecular Biology, Entomology, and Plant Pathology, Mississippi State University, Dorman Hall 402, Box 9655, Mississippi State, MS 39762. 2Department of Plant and Soil Sciences, 117 Dorman Hall, Mississippi State University, Starkville, MS 39759.3Arkansas Agricultural Experiment Station, AFLS 212, University of Arkansas, Fayetteville, AR 72701. 4Great Smoky Mountains National Park, 107 Park Headquarters Road, Gatlinburg, TN 37738. *Corresponding author - rbaird@plantpath.msstate.edu. Manuscript Editor: Tim Lindblom Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 2 understory of Rhododendron maximum L. (Rosebay Rhododendron; hereafter Rhododendron) (Boettcher and Kalisz 1990, Madden et al. 2004). However, Eastern Hemlock is currently being killed at very high rates by an exotic insect, Adelges tsugae Annand (Hemlock Woolly Adelgid). Loss of this tree species from the overstory is especially significant along riparian zones, increasing the importance of secondary canopies of remaining smaller-sized woody trees such as Rhododendrons, which shelter the forest floor from direct sunlight and subsequent higher soil temperatures. Rhododendron is an important subcanopy plant in the southern Appalachian Mountains and is considered a keystone species for those forest ecosystems (Pickering et al. 2003). Rhododendron provides valuable winter escape and cover for numerous forest animals and has possible competitive inhibitory effects to other vegetation. In the late 1980s, large areas of Rhododendron were found dying throughout the southern Appalachians. In particular, 14 major damaged locations were identified by National Park Service personnel as of 2006 (K. Johnson, pers. observ.). In 1993, Ownley (1993, 1994) conducted an initial study on Rhododendron dieback to quantify the amount of damage and identify putative causal factors. The study concentrated on above- and belowground biotic and abiotic factors at two known locations. No microbial agents were isolated or identified, but a viable tree-health rating was developed. With the exception of that initial six-month study, no other research on this problem was conducted. With no biotic or abiotic causes identified, a general survey (Baird et al. 2013) and intensive study evaluating other critical site parameters and soil microbes was considered the next phase of research. Several microbes are known to be damaging to landscape and forest Rhododendrons. Agrobacterium tumefaciens [E.F. Sm. & Town.] Conn. causes a crown gall on roots of woody plants (Coyier and Roane 1988, Sinclair et al. 1993). However, few species are significantly damaged and, for Rhododendron spp., the disease seems to be minor. The bacterium is widespread and is known to predispose the plants to secondary invasion by Armillaria spp. (Moore 1988, Sinclair et al. 1993). Much like bacteria, there has been little research about viruses which may impact Rhododendrons. More recently, one species identified as Virus A (RhVA) has been commonly found in the southern Appalachian Mountains and appears to cluster with Southern Tomato Virus based on molecular data (Sabanadzovic et al. 2010). Several other noted plant pathogens are reported on Rhododendrons, including Botryosphaeria dieback, caused by Botryosphaeria dothidea (Moug.) Ces. & De Not. This pathogen is considered an opportunistic fungus that attacks predisposed hosts causing leaf blight, spot, and stem dieback, but rarely causes death of the whole Rhododendron plant (Jones 1988, Sinclair et al. 1993). Ownley (1993, 1994) did not observe this pathogen in the dieback plots within GRSM. Phytophthora spp. are among the most damaging and economically important diseases of ornamental crops in the southeastern United States (Benson and Hoitink 1988, Ferguson and Jeffers 1999, Hoitink et al. 1988, Linderman 1988). Root rot from Phytophthora cinnamomi Rands usually is found on younger Ericaceous plants. Nematodes, the most plentiful metazoans, are pseudocoelomate worm-like animals and have routinely been found in GRSM (Bernard 2005). Some species have Southeastern Naturalist 3 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 been associated with dieback and decline of Rhododendrons (Benson and Barker 1988). Problems linked to nematodes observed on Rhododendron have usually been restricted to warmer regional areas of sandy soil [REF?]; however, they are a natural part of the forest ecosystem and may thrive under aberrant conditions (Bernard 2005). Deficiencies or imbalances of soil nutrients in plants or trees can cause stress and allow organisms to impact forest health. An imbalance of a specific nutrient level can influence the availability of other nutrients depending upon pH levels in the soil (Fink 1999). Therefore, mineral nutrient excesses and deficiencies can be a major factor in the health of Rhododendron and other forest trees. Subsequently, mycorrhizal associations may be reduced while enhancing secondary invaders resulting in even further reduction of nutrient uptake ability of the host (Fink 1999). Thus, nutrient availability as impacted by pH may play a major role in overall forest health and in problems such as Rhododendron dieback. Many biotic and abiotic soil parameters are known to affect plant health; therefore, the objectives of this study were first to evaluate select microbial, nematode, plant, and site parameters at two locations where rhododendron dieback has been reported in the Great Smoky Mountains National Park (GRSM) and Nantahala National Forest (NNF); second, to obtain baseline data on the presence and/or associations of potential soilborne pathogens or other microbes from roots and soil rhizosphere; and third, to define and correlate dieback/decline within Rhododendron stands compared with data obtained in the first objective. The results from this study will attempt to determine the cause(s) of Rhododendron dieback in the southern Appalachian Mountains. Materials and Methods Plot establishment and data collection Two sites with large areas of declining or dying Rhododendron plants were evaluated, one at Albert Mountain, NNF (35°05.257'N, 83°48.024'W) and the other in the Laurel Falls area of GRSM (35°68.159'N, 83°60.080'W). At Albert Mountain, the soil consisted of a Burton sandy loam (fine-loamy, isotic, frigid Typic Humudept) and a Craggery sandy loam (loamy isotic, frigid Lithic Humudept) at an elevation of 1541.4 m. At Laurel Falls, the soil was a Snowbird loam-sandy clay loam (fine-loamy, mixed, active, mesic Humic Hapludult) at an elevation of 945.0 m. These study sites were also used in a general survey of dieback locations, using one large plot 100 × 100 m, each in GRSM and NNF (Baird et al. 2013). Within each of the two dieback areas, eight plots 20 × 20 m in size (400 m2) were selected based on health ratings (see below) of the Rhododendron plants and consisted of four dieback and four control (healthy) plots (Table 1). To determine Rhododendron dieback severity levels within each of the plots, all Rhododendron plants were surveyed and R values calculated as shown below using the scale in Table 2 (Ownley 1994). Disease ratings were obtained during May–June in 2006. Plots with mean disease rating values of >5 were considered dieback plots and those with values of less than 2 were considered healthy/control plots. Besides shrub Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 4 dieback ratings within each plot, We collected data on site evaluation parameters such as aspect, slope, pH, and soil nutrients. Nematode levels in plots were measured and reported as concentrations per 473-cc soil samples. All methods for site parameters followed previous methods (Baird et al. 2013). We collected information on potential plant pathogen associations from root and soil samples from each plot using traditional isolation methods and molecular environmental sampling data (discussed below). In May 2006, we examined all trees for signs of dieback in the plots at both intensively sampled sites. To ensure the entire plot was sampled, four transects were established extending 3 m from the midpoint (plot center) to within 10 m of each corner. Each Rhododendron plant >3 cm diameter at breast height (dbh) was evaluated using the rating scale in Table 2 (Ownley 1994). This scale has been adopted Table 1. Plot locations (GPS coordinatesA) of the two rhododendron dieback/control study areas associated with decline of R. maxium (Rosebay Rhododendron) trees in Great Smoky Mountains National Park (GRSM) and Nantahala National Forest. Laurel Falls, GRSM Dieback Plots (RDLF01BB) (RDLF03B) (RDLF04B) (RDLF08B) 35°68.159'N 35°68.147'N 35°68.142'N 35°68.119'N 83°60.080'W 83°60.058'W 83°60.115'W 83°60.072'W Healthy Plots (RDLF01G) (RDLF02G) (RDLF04G) (RDLF05G) 35°68.150'N 35°68.156'N 35°68.133'N 35°68.113'N 83°60.236'N 83°60.202'W 83°59.989'W 83°59.996'W Albert Mountain, Nantahala N.F. Dieback Plots (RDAM02B) (RDAM03B) (RDAM06B) (RDAM08B) 35°05.368'N 35°05.354'N 35°05.378'N 35°05.404'N 83°13.669'W 83°47.931'W 83°47.920'W 83°47.898'W Healthy Plots (RDAM03G) (RDAM09G) (RDAM16G) (RDAM17G) 35°05.567'N 35°05.492'N 35°05.533'N 35°05.504'N 83°47.798'W 83°47.759'W 83°47.842'W 83°47.800'W AUTM NAD27 CONUS. BB at the of each plot number = dieback plots; G = healthy plots. Table 2. Rating scale for evaluating or estimating rhododendron dieback within R. maximum stands in Great Smoky Mountains and Nantahala National Forest. # of dead twigs # of whorlsA WH-valueA % Chlorosis CH-value and branches TW-value ≥4 whorls 1 0 1 <5% 1 3–4 whorls 3 0–10% 2 5–20% 4 2–3 whorls 4 10–25% 3 >20% 5 1–2 whorls 5 >25% 5 Recently deadB 6 DeadC 7 AThe values assigned to WH range from 1 to 7, with 1 being healthy and 7 being completely dead. BRecent dead refers to dead plants prior to leaf abscission. CDead refers to dead plants with no leaves attached. Southeastern Naturalist 5 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 by USDA/Park Service (Glenn Taylor, pers. observ.). The estimated leaf whorls per branch (WH), percentage of leaves with chlorosis (CH), and percentage of dead twigs and branches (TW) were assigned specific values (Table 2). These values then were analyzed using the rhododendron dieback rating modified formula (Ownley 1994). We excluded 2 variables (canopy closure [CA] and other [OT]) from the original formula to make the analysis more robust: R_Calc = 1.00/3(WH + CH + TW) or R= (WH + CH + TW)/3 R values can range from 1 to 7 for WH, with 1 defined as a healthy plant and 6 or 7 as dead before or after leaf abscission, respectively. The R value indicates the severity of dieback for all trees within each plot and across each location. Soil samples for nematode analysis were collected annually during August in 2006 through 2008. All nutrient samples were collected in May for each of the three years. From the NE to SW sides of each plot, sampling points were designated along a horizontal axis at 5, 10, and 15 m (three sampling points ) and along a vertical axis from NW to SE at 5, 10 and 15 m (three sampling points) and at 4, 8, 12, and 16 m (four sampling points). A total of 10 soil samples for pH, nutrient analysis and nematodes were taken from each plot. The 10 soil samples were combined, mixed, and stored at 20 °C, until submitted for processing. All soil samples (2 L) were collected with small hand shovels that were cleaned between plots. For nutrient and pH analyses, published methods were used (Cox 2001, MSU 2004). Methods for nematode extraction from the soil followed those by Baker (1978). In addition, juvenile, vermiform, and cyst stages of nematodes were isolated using the North Carolina-style semi-automatic elutriator and sugar centrifugation methods (C. Balbalian, Diagnotic Lab., Entomology and Plant Pathology Department, MS, pers comm) (MSU 2005). After the cyst counts were obtained, all cysts were placed onto a Baermann funnel (Baermann funnel method) for five days, and the number of larvae that hatched was counted. Identification of potential pathogens Samples of the Rhododendron rhizosphere (root/soil zone) were collected on 10–11 August 2007. A total of 10 plants per plot were randomly selected, numbered, and tagged based on their individual R value and plot number. Those plants located in the severe plots had R ratings of >4 to less than 6. Those in the healthy plots had ratings of less than 3. We collected a 100-cm-length sample of secondary/feeder roots with attached soil from the north side and within the dripline of each plant canopy. The roots were collected from the 0 to 10 cm soil horizon (A–B horizons). The samples were placed into a 3.8-L Ziploc® plastic bag, transferred to refrigeration (4 °C), and processed within 7 days of collection. We further divided the 100-cm secondary/ feeder root samples into two equal sections of 50 cm, and randomly selected each for (1) cultural isolation and identification and (2) molecular identification. For the cultural-isolation studies, we used a direct plating method for each of the randomly selected 50-cm secondary/feeder root samples (Pearson and Read 1973). We cleaned the roots in a 2.0-mm mesh sieve (No. 10) under gently running tap water for 10 min to remove soil and adhering debris (Dighton and Harrison 1983, Goodman et al. 1996). The sample was surface disinfected in 0.525% (w/v) aqueous Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 6 NaOCl for 2 min. (Villarroel et al. 2004). Using a scalpel, we excised two secondary and two feeder roots (equally spaced) at 1 cm each in length and plated them onto malt extract agar (MEA) and potato dextrose agar (PDA- Difco®, Detroit, MI) amended with 30 mg/l of streptomycin sulfate and 50-mg/l of chlortetracycline as bacteriostatic agents (Pearson and Read 1973, Tuite 1969, Villarroel et al. 2004) in 100- × 15-mm Petri dishes. We then incubated the plates at room temperature with a photoperiod of 12 h and examined them at regular intervals for 4 days (Pearson and Read 1973). All fungi growing from the root parts were subcultured onto MEA and maintained as described above. We stored representative fungal collections at -80 °C in a 1.2-ml autoclaved cryogenic vial (Corning, Acton, MA), containing 15% glycerol for later characterizations. To prepare the isolates for identification using cultural morphologies, we placed mycelium plugs on plates of PDA (Villarroel et al. 2004) and incubated them at room temperature with a photoperiod of 12 h (Pearson and Read 1973). We subcultured the resulting colonies for 14 days and used morphological characteristics to identify the fungi as per Barnett and Hunter (1998), Ellis (1971), and Sutton (1980). For initial screening, single-spored isolates of Fusarium spp. were evaluated using carnation-leaf agar (Toussoun and Nelson 1976). We prepared the carnation leaf agar medium using four discs of prepackaged sterile irradiated carnation leaves (Fusarium Research Center, Buckout Laboratory, Pennsylvania State University, State College, PA). Molecular identification of fungi: We subdivided the 50-cm root samples (10 per plot) into 8 equally spaced 6.25-cm segments. From the center of each segment, 1.0-cm pieces were excised and placed into 2X CTAB for storage and later identification, using molecular procedures outlined below. We isolated fungal DNA using methods previously discussed (Baird et al. 2010). We stored all genomic DNA at -80 °C. Molecular methods follow those commonly used for fungal identification (Mata et al. 2007, O’Brien et al. 2005). The ITS region (ITS 1, 5.8 rDNA gene, ITS 2) was amplified using primer pairs ITS1F and ITS4 (Gardes and Bruns 1993, White et al. 1990) as well as cloning and sequencing. (Baird et al. 2010, Lickey et al. 2007). We aligned all sequences, determined percent similarities (Lasergene-SeqPro, Madison, WI, USA), and conducted a GenBank (NCBI) BLAST search to determine their identities. Following the completion of the research, all sequence data will be deposited in Genbank (NCBI). For Fusarium spp., all procedures followed those by Geiser et al. (2004). To determine if Oomycota root organisms were possibly involved in the decline, we collected soil samples adjacent (≤ 5.0 cm) to roots of 10 plants per plot where the 100-cm length of secondary/feeder roots was collected for fungal mycoflora study as discussed previously. The soil subsamples within each plot were pooled and mixed in a disinfected 19-L polypropylene bucket to form a composite sample of approximately 1.0-L. Each composite sample was placed into a 3.8-L Ziploc® plastic bag and stored at 4 °C. Within 7 days, we assayed soil samples for Pythiaceous stramenopiles using a baiting bioassay identical to previously designed methods Southeastern Naturalist 7 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 (Ferguson and Jeffers 1999, Jeffers and Aldwinckle 1987, Jeffers and Martin 1986). Using a baiting bioassay, we assayed six 100-ml aliquots from each screened soil sample: three aliquots were baited fresh and three aliquots were baited after soil had been air-dried as per previous studies. The rhododendron leaves collected at the field sites used for baiting were plated onto P5ARPH-V8 selective medium to recover Phytophthora spp. (Ferguson and Jeffers 1999, Jeffers and Martin 1986), and we placed pear cubes on P5ARP-V8 selective medium to recover Pythium spp. We transferred representative isolates onto fresh P5ARPH-V8 or P5ARP-V8 medium. We grew isolates of putative Pythiaceous stramenopiles on clarified V8A medium in 8.0-ml glass vials at 20 to 25 °C for 10 to 14 days, and then stored the vials at 12 to 15 °C. Statistical analysis The experimental plots were assigned within specific locations based on treatment requirements at each site. The two treatments were diseased and healthy (control), with four replicates of each. The measure of fungal species diversity included species richness (SR) and Shannon diversity index (H'; Van Dyke 2003), which was calculated as: H' = -Σ (pi ln pi), (i = 1,2,3,….S), where pi is the proportion of the total population and S is total number of species. Published methods were used to calculate coefficient of community (CC; Van Dyke 2003) and evenness (J; Stephenson 1989, Stephenson et al. 2004). Stephenson (1989) provides a thorough description of all formulas for these indices. Within each site, there were two treatments (dieback and control), with four subplots within each treatment (Baird et al. 2013). We analyzed data as a series of combined experiments (combined across sites) using the GLM procedure of SAS (SAS Institute 1999) and separated means using Fisher’s protected least significant difference (LSD). We pooled data within each location (e.g., Albert Mountain and Laurel Falls) and used stepwise multiple regression analysis to evaluate the effect of the measured variables on dieback rating (R_Calc). Results and Discussion The root tissues sequencing data resulted in 106 species of fungi, pooled across locations. These included primarily anamorphic Ascomycota, Basidiomycota, and very few species from other phyla (Table 3). These results may be an indication of the selectivity of the ITS primers used during the investigation. However, the primers used in this investigation were previously reported to have a broad range of fungal selectivity (Borneman and Hartin 2000). The most common fungus obtained from the sequence data was identified as unknown Basidiomycota (sp. 1) but later identified as Mycena silvae-nigrae Maas Geest. & Schwöbel from GenBank database (Table 3). Mycena silvae-nigrae occurred commonly at both locations, and the population was generally uniform across the dieback and healthy plots, except that percent frequency was significantly lower Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 8 Table 3. Mean percent occurrenceA of fungi from sequence/fungal clone data of R. maxium (Rosebay Rhododendron) roots from two dieback locationsB in the southern Appalachian Mountains. Species epithets provided here and in Tables 5 and 6 are based on sequence identification from NCBI GenBank library of accessions, which does not provide authority information. Albert Mountain Laurel Falls Taxa/GenBank accession number Healthy Dieback Healthy Dieback Ascomycota Capnodiales /JQ272378 0.0 0.0 7.6 0.0 Clavicepitaceae /JQ272381 0.0 0.0 <1.0 <1.0 Cenococcum geophilum /JQ272423 4.7 0.0 0.0 0.0 Dermateaceae /JQ272373 <1.0 0.0 1.6 0.0 Elaphomyces sp. 1 /JQ272414 1.8 0.0 0.0 0.0 Glomeromycete 1 /JQ272369 0.0 0.0 6.0 1.1 Helotiales 1 /JQ272327 0.0 0.0 0.0 3.6 Helotiales 2 /JQ272329 0.0 0.0 0.0 <1.0 Helotiales 3 /JQ272334 0.0 1.0 0.0 3.9 Helotiales 4 /JQ272371 4.7 0.0 <1.0 0.0 Helotiales 5 /JQ272385 0.0 0.0 <1.0 0.0 Helotiaceae 1 /JQ272370 0.0 0.0 5.5 0.0 Helotiaceae 2 /JQ272393 3.1 <1.0 0.0 0.0 Helotiaceae 3 /JQ272432 <1.0 0.0 0.0 0.0 Hymenoscyphus sp. 1 /JQ272341 0.0 0.0 0.0 2.8 Herpotrichiellaceae 1 /JQ272383 0.0 0.0 1.0 0.0 Hyaloscyphaceae /JQ272392 0.0 11.1 0.0 0.0 Hypocreales /JQ272387 0.0 0.0 <1.0 0.0 Leotiomyceta 1 /JQ272338 0.0 0.0 0.0 4.2 Leotiomyceta 2 /JQ272350 0.0 0.0 0.0 <1.0 Leotiomyceta 3 /JQ272351 0.0 0.0 0.0 <1.0 Leotiomyceta 4 /JQ272352 0.0 0.0 0.0 <1.0 Leotiomycetes 1 /JQ272356 0.0 0.0 2.9 <1.0 Lecanoromycetidae /JQ272389 0.0 0.0 0.0 <1.0 Magnaporthaceae 1 /JQ272354 0.0 0.0 <1.0 <1.0 Magnaporthaceae 2 /JQ272424 <1.0 0.0 0.0 0.0 Meliniomyces variabilis /JQ272408 0.0 1.5 0.0 0.0 Oidiodendron sp. 1 /JQ272359 16.4 0.0 2.8 0.0 Ophiostomaceae /JQ272396 0.0 1.8 0.0 0.0 Phialocephala sp. 1 /JQ272328 <1.0 0.0 2.9 2.8 Penicillium spinulosum /JQ272372 0.0 0.0 <1.0 0.0 Penicillium waksmansii /JQ272380 0.0 0.0 <1.0 0.0 Phialocephala fortinii /JQ272400 3.1 <1.0 0.0 0.0 Pezizomycotina /JQ272412 <1.0 0.0 0.0 0.0 Rhizoscyphus sp. 1 /JQ272361 11.5 <1.0 <1.0 <1.0 Rhizoscyphus sp. 2 /JQ272427 2.1 0.0 0.0 0.0 Rhizoscyphus ericae /JQ272407 <1.0 0.0 0.0 0.0 Rhytismataceae /JQ272405 <1.0 <1.0 0.0 0.0 Sordariomycetes 1 /JQ272360 2.6 3.3 0.0 <1.0 Sordariomycetes 2 /JQ272429 2.1 0.0 0.0 0.0 Trichoderma asperellum /JQ272391 0.0 0.0 3.4 0.0 Verrucariales /JQ272347 1.8 0.0 1.8 1.9 Unknown Acomycete 1 /JQ272331 0.0 0.0 <1.0 6.1 Unknown Ascomycete 2 /JQ272339 0.0 0.0 0.0 2.2 Unknown Ascomycete 3 /JQ272340 0.0 0.0 0.0 7.2 Southeastern Naturalist 9 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 Table 3, continued. Albert Mountain Laurel Falls Taxa/GenBank accession number Healthy Dieback Healthy Dieback Unknown Ascomycete 4 /JQ272341 0.0 0.0 0.0 1.9 Unknown Ascomycete 5 /JQ272357 0.0 <1.0 1.0 <1.0 Unknown Ascomycete 6 /JQ272358 0.0 0.0 <1.0 1.1 Unknown Ascomycete 7 /JQ272384 1.8 0.0 2.9 0.0 Unknown Ascomycete 8 /JQ272388 0.0 0.0 3.4 0.0 Unknown Ascomycete 9 /JQ272398 1.8 1.8 0.0 0.0 Unknown Ascomycete 10 /JQ272422 <1.0 0.0 0.0 0.0 Unknown Ascomycete 11 /JQ272431 2.1 0.0 0.0 0.0 Basidiomycota Agaricales 1 /JQ272363 0.0 0.0 2.9 1.9 Amanita sp. 1 /JQ272374 0.0 0.0 1.0 0.0 Amanita sp. 2 /JQ272418 <1.0 0.0 0.0 0.0 Clavulina cinerea /JQ272409 0.0 7.1 0.0 0.0 Cortinarius firmus /JQ272395 0.0 2.0 0.0 0.0 Cortinarius sp. 1 /JQ272415 <1.0 0.0 0.0 0.0 Cortinarius sp. 2 /JQ272416 0.0 0.0 <1.0 0.0 Cuphophyllus lacmus /JQ272404 0.0 <1.0 0.0 0.0 Cystofilobasidium infirmo-miniatum /JQ272390 0.0 0.0 <1.0 0.0 Galerina fibrillosa 1 /JQ272325 0.0 0.0 0.0 21.4 Galerina sp. 1 /JQ272382 0.0 2.8 2.6 0.0 Gymnopus sp. 1 /JQ272362 0.0 0.0 6.8 <1.0 Lactarius imperceptus /JQ272401 <1.0 7.8 0.0 0.0 Lactarius sp. 1 /JQ272335 0.0 0.0 0.0 <1.0 Lactarius sp. 2 /JQ272344 0.0 0.0 0.0 6.2 Leucosporidium sp. 1 /JQ272411 0.0 1.5 0.0 0.0 Marasmius scorodonius /JQ272364 0.0 0.0 1.6 5.3 Mycena sp. 1 /JQ272379 0.0 0.0 1.0 0.0 Nolanea sp. 1 /JQ272420 <1.0 0.0 0.0 0.0 Russulaceae /JQ272413 0.0 1.0 0.0 0.0 Russula atropurpurea /JQ272366 0.0 0.0 0.0 1.1 Russula granulata /JQ272365 0.0 0.0 9.4 0.0 Russula xerampalina /JQ272428 3.9 0.0 0.0 0.0 Russula sp. 1 /JQ272330 0.0 0.0 0.0 2.2 Russula sp. 2 /JQ272402 <1.0 2.3 0.0 0.0 Russula sp. 3 /JQ272403 1.8 1.8 0.0 0.0 Russula sp. 4 /JQ272421 <1.0 0.0 0.0 0.0 Russula sp. 5 /JQ272425 1.3 0.0 0.0 0.0 Sebacinales 1 /JQ272332 0.0 0.0 0.0 3.6 Sebacina sp. 1 /JQ272410 2.1 0.0 0.0 0.0 Sebacina sp. 2 /JQ272430 <1.0 0.0 0.0 0.0 Thelephoraceae /JQ272375 0.0 <1.0 2.0 0.0 Tormentella sublilacina /JQ272367 0.0 0.0 0.0 <1.0 Tormentella sp. 1 /JQ272406 2.3 2.8 0.0 0.0 Tormentella sp. 2 /JQ272433 <1.0 0.0 0.0 0.0 Tricholomataceae /JQ272419 1.6 0.0 0.0 0.0 Xenasmataceae /JQ272394 2.1 12.9 0.0 0.0 Unknown Basidiomycota 1 /JQ272326 (Mycena silvae-nigrae) 10.2 19.7 15.9 1.7 Unknown Basidiomycota 2 /JQ272333 0.0 0.0 0.0 2.4 Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 10 from the Laurel Falls dieback plots. Also, Galerina fibrillosa A.H. Smith frequencies were significantly greater at Laurel Falls, and we only found it in the dieback plots. This species is considered to be a saprophyte and is not associated with Rhododendron dieback. Other common species were Ascomycota such as Rhizoscyphus sp. 1 and Oidiodendron sp.1, occurring at frequencies of 12.7% and 19.2%, respectively, from healthy plots at Albert. Occurrences in other plots were minimal for these two species. Rhizoscyphus spp., especially Rhizoscyphus ericae (D.J. Read) W.H. Zhuang & Korf., were reported to form mycorrhizal associations with Ericaceous plant species such as Rhododendron (Grelet et al. 2010). This fungal species forms a complex on roots of Ericaceous shrubs and ectomycorrhizal-forming trees. Other basidiomycota, such as Lactarius imperceptus Beardslee & Burl. (ectomycorrhizae) and Gymnopus sp.1 (saprophyte), were locally abundant, but we noted no trends; isolation frequencies were 7.8% from dieback plots at Albert Mountain and 6.8% from healthy plots at Laurel Falls. We compared species richness, diversity, evenness, and community coefficients from the root sequence data for the different fungal clones (Table 4). Species richness values were significantly lower at Albert Mountain in dieback plots than in healthy plots and were greater where Rhododendron and the lesser vegetation was in abundance. The open understory may have been drier and had higher temperatures due to the increased sunlight reaching the forest floor at Albert Mountain, resulting in lower fungal populations occurring in rhizosphere layers there. At Laurel Falls, species richness was similar between dieback and healthy plots. Diversity of the fungi at Albert Mountain was also significantly lower in dieback plots compared to healthy plots. No other differences were noted between locations and Table 3, continued. Albert Mountain Laurel Falls Taxa/GenBank accession number Healthy Dieback Healthy Dieback Unknown Basidiomycota 3 /JQ272336 0.0 0.0 0.0 7.2 Unknown Basidiomycota 4 /JQ272345 0.0 0.0 0.0 <1.0 Unknown Basidiomycota 5 /JQ272368 0.0 0.0 0.0 3.9 Unknown Basidiomycota 6 /JQ272376 <1.0 0.0 <1.0 0.0 Unknown Basidiomycota 7 /JQ272377 0.0 0.0 <1.0 0.0 Zygomycotina Mortierellales /JQ272348 0.0 0.0 0.0 <1.0 Unknown Fungi 1 /JQ272342 <1.0 0.0 0.0 <1.0 Unknown Fungi 2 /JQ272343 0.0 0.0 0.0 <1.0 Unknown Fungi 3 /JQ272349 0.0 <1.0 0.0 0.0 Unknown Fungi 4 /JQ272386 0.0 0.0 0.0 <1.0 Unknown Fungi 5 /JQ272417 <1.0 0.0 0.0 0.0 Unknown Fungi 6 /JQ272426 1.0 0.0 0.0 0.0 Unidentified /JQ272337+JQ272397+JQ272399 <1.0 9.9 8.1 3.3 APercent isolation frequency based on 2 locations × 8 Rhododendron plots (4 dieback + 4 healthy control) × 1 genomic DNA sample/plot (10 plants/plot with eight 1-cm pieces removed/plant, all 80 pieces pooled for extraction/plot) × 100 fungal clones or 1600 clones. BAlbert Mountain is located in Nantahala National Forest, and Laurel Falls is in the Great Smoky Mountains National Park. Southeastern Naturalist 11 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 treatments. Also, evenness across locations and treatments were similar, with all having relative high abundance values (E = 0.86 overall). We obtained coefficient of community values for fungal taxa to compare data for dieback and healthy plots at each location and across pooled locations (Table 4). The CC values indicated taxa were most common within locations. Similarity of taxa across locations and between healthy and dieback plots within locations ranged from 25 to 42%. Laurel Falls dieback and healthy plots had the highest similarity of taxa at 0.40% CC, and the lowest similarity occurred between both locations when comparing between healthy and dieback or just dieback plots for locations. Overall, the CC values for occurrences of taxa were similar within locations, but varied much more between locations. We isolated a total of 31 species of fungi from the roots collected at both locations (Table 5). The most common species isolated could only be identified as Helotiales (Ascomycota) based on sequence data. Furthermore, the cultures did not sporulate, making further identification impossible. Morphologically, the Helotiales culture was similar to Rhizocyphus sp., but the sequences indicated 90% or greater similarity. Walker et al. (2011) reported that members of the Helotiales such as R. ericae complex form mycorrhizal associations with Ericaceae. In that study, members of Helotiales were considered important in ericoid mycorrhizal associations. Isolation of the Helotiales 1–5 spp. from roots was somewhat common across locations and treatments. Ilyonectria radicicola (Gerlach & L. Nilsson) P. Chaverri & C. Salgado, the sexual stage of Cylindrocarpon destructans (Zinssm.) Scholten., noted as an important root pathogen of forest trees, herbaceous and wood plants, and nursery seedlings (Levy et al. 2007), and was common at both locations and treatments in this study. Isolation frequencies across pooled location data and treatments ranged from 10.1% to 30.7% for I. radicicola. Another common species, Hypocrea viridescens Jaklitsch and Samuels, was isolated from both locations and across treatments and ranged from 8.2% to 21.0% isolation frequencies. Five Table 4. Species richness (n), diversity (H'), and evenness (E) for fungi identified from R. maxium (Rosebay Rhododendron) root samples obtained from two locations in the southern Appalachian Mountains. D = dieback plot, H = healthy plot. Coefficient of community Location n H' J Location AB B C CC Albert D 27 BA 2.72 B 0.83 A Albert D vs H 27 45 13 0.36 Albert H 45 A 3.19 A 0.84 A Albert D vs Laurel D 27 43 6 0.17 Laurel D 43 A 3.14 AB 0.83 A Albert D vs Laurel H 27 37 5 0.16 Laurel H 37 AB 3.08 AB 0.85 A Albert H vs Laurel D 45 43 7 0.16 LSD (10) (0.46) (0.10) Albert H vs. Laurel H 45 37 10 0.24 Total comparison 106 4.03 0.86 Laurel D vs. Laurel H 45 37 16 0.40 AMeans within columns followed by same letter are not significantly different (P ≤ 0.05) according to the LSD. BA, B, and C = number of species identified at Albert Mountain, Laurel Falls, or both locations respectively; CC = (2*C) / (A + B). Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 12 species of Penicillium and four of Trichoderma were identified across both locations but at low frequencies. Species of Hypocrea (anamorph-Trichoderma spp. primarily) are generally considered saprophytes or hyperparasites of other fungi (Rossman et al. 1999). Four species of Basidiomycota and three of Zygomycota were identified across locations. No differences between species richness, diversity, Table 5. Mean percent occurrence of fungi from isolation data of R. maxium (Rosebay Rhododendron) roots from two locations in the southern Appalachian Mountains. Albert Mountain, Laurel Falls, Nantahala National Forest Great Smoky Mountains Taxa Healthy Dieback Healthy Dieback Ascomycota Chaunopycnis alba /JQ272438 1.7 0.0 2.5 3.4 Cryptosporiopsis eraceae /JQ272452 2.4 3.4 3.1 1.1 Dendrosporium sp. 1 /JQ272461 0.0 <1.0 0.0 0.0 Helotiales /JQ272459 31.6 30.7 32.7 23.4 Hypocrea lixii /JQ272437 5.3 6.1 6.9 9.7 Hypocrea pachybasidiodes /JQ272449 1.2 2.3 0.0 2.3 Hypocrea sinosa /JQ272463 0.0 0.0 <1.0 0.0 Hypocrea viridescens /JQ272436 15.8 21.0 8.2 8.6 Lachnum virgineum /JQ272454 1.1 1.2 1.3 <1.0 Ilyonectria radicicola /JQ272460 31.6 30.7 10.1 14.9 Penicillium citreonigrum /JQ272434 1.7 0.0 2.5 3.4 Penicillium corylophilum /JQ272455 0.0 1.7 0.0 0.0 Penicillium janthinellum /JQ272458 1.8 1.1 3.8 1.7 Penicillium spinulosum /JQ272447 1.2 1.7 1.3 <1.0 Ponchonia bulbillosa /JQ272440 2.4 2.3 3.8 1.1 Ponchonia sp. 1 /JQ272439 1.2 <1.0 <1.0 1.1 Pezizomycotina 2 /JQ272453 0.0 1.7 0.0 0.0 Phialocephala fortinii /JQ272457 1.8 1.7 2.5 1.7 Phialocephala sp. 2 /JQ272456 2.9 3.4 4.4 3.4 Thysanophora penicillioides /JQ272462 <10.0 0.0 <1.0 1.1 Trichoderma koningiopsis /JQ272435 5.8 6.3 3.8 5.1 Trichoderma pubescens /JQ272444 0.0 <1.0 1.3 0.0 Trichoderma viride /JQ272443 1.2 0.0 0.0 <1.0 Unknown Ascomycete 1 /JQ272442 2.9 0.0 2.5 <1.0 Basidiomycota Gymnopilus sp.1 (penetrans ?) /JQ272441 2.4 1.1 1.3 <1.0 Mycena sp. 2 /JQ272464 3.5 <1.0 2.5 4.0 Tremellomyces 1 /JQ272445 1.8 1.1 1.3 <1.0 Basidiomycete 1 /JQ272446 <1.0 1.1 0.0 1.1 Zygomycota Mortierella macrocystis /JQ272448 1.8 0.0 <1.0 <1.0 Unknown Fungi 1 /JQ272450 1.2 0.0 1.3 <1.0 Unknown Fungi 2 /JQ272451 <1.0 0.0 1.3 0.0 APercent isolation frequency based on 2 locations × 8 plots (4 healthy and 4 dieback) × 40 root pieces (2 feeder roots 1 cm long and 2 secondary roots 1 cm long for 10 plants) × 2 media (4 pieces/plate) × 5 isolates/plate = 6400 total isolations; sometimes less than 5 isolates/plate were found; all data confirmed by morphological features and sequence data. Southeastern Naturalist 13 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 evenness, or community coefficient occurred for root isolations of the 32 fungi cultured using artificial media. The selective media PARPH-V8 and PARP-V8, which had been thought to be specific for Oomycota (e.g., Phytophthora and Pythium), have been reported to commonly allow Mortierella spp. (Zygomycota) to be cultured (S. Jeffers, Clemson University, Clemson, SC, pers. comm.). In the baiting study, 12 species of fungi were isolated and four species of fungi were commonly observed from the PARPV8 media. These included Mortierella humalis Linnemann ex W. Gams, Mortierella macrocystis W. Gams, Mortierella sp. 1, and Mortierella sp. 2 (Table 6). These four species had significantly greater isolation frequencies than any other fungi. We isolated no species of Phytophthora or Pythium using this medium, indicating that there is no association of Oomycota to Rhododendron dieback. This finding follows preliminary reports that no pathogenic Phytophthora spp. were isolated or sequenced when Rhododendron roots from dieback areas of western North Carolina were evaluated (K. Ivors, North Carolina State University, Raleigh, NC, and S. Oak, USDA Forest Service, Ashville, NC, pers. comm.). In addition, temperatures during 30 days prior to sampling averaged 18 °C, which is sufficient for Oomycota to occur on the roots if present. None of the four Mortierella spp. are considered pathogens of woody ornamentals or Rhododendron, but are common saprophytic soil inhabitants (Watanabe 1994). Table 6. Mean percentA occurrence of rhizosphere organisms isolated from roots on PARPB media at two rhododendron dieback sites in the southern Appalachian Mountains. Albert Mountan, Laurel Falls, Nantahala National Forest Great Smoky Mountains Taxa Healthy Dieback Healthy Dieback Ascomycota Fusarium proliferatum /JQ272470 5.6 3.7 5.8 1.4 Penicillium corylophilum /JQ272469 5.6 11.1 3.5 8.9 Trichoderma asperellum /JQ272474 4.7 7.8 5.4 11.7 Trichoderma hamatum /JQ272476 4.9 6.6 4.7 7.0 Trichoderma tomentosum /JQ272475 5.1 4.9 3.5 4.2 Venturiaceae 1 /JQ272465 4.5 3.2 16.3 5.6 Zygomycotina Mortierella humalis /JQ272471 27.1 13.2 11.3 13.6 Mortierella macrocytis /JQ272473 5.1 5.8 18.7 15.0 Mortierella sp. 1 /JQ272467 8.9 20.6 9.7 14.0 Mortierella sp. 2 /JQ272468 16.8 4.9 8.9 5.6 Unknown Taxa Unknown sp. 1 /JQ272466 16.1 16.0 11.3 7.5 Unknown sp. 2 /JQ272472 1.9 2.9 4.7 5.6 APercent frequency is based on number of isolated obtained from soil baiting procedure with Albert Mountain having 212 cultures in healthy plots and 245 in dieback plots; Laurel Falls had 248 cultures in healthy plots and 208 in the dieback plots. BPARP = special medium for isolation for Oomycetes and includes pimaricin + ampicillin + rifampicin + pentachloronitrobenzene. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 14 Species richness values for 12 species were similar between the two locations and between the dieback and healthy plots (data not shown). However, species diversity and evenness at Laurel Falls (2.40 and 0.72, respectively) were significantly greater than at Albert Mountain (2.30 and 0.92, respectively). As stated above, none of the organisms isolated were pathogenic stramenopile species, but the data only reflected saprophytic fungal populations. We collected soil samples to determine the potential importance of nematode species in rhododendron dieback. The most common species were Criconemella xenoplax Raski (ring), Parathrichodorus minor (Colbran) Siddiqi (stubby root), and Meloidogyne sp. (root knot) across years and treatments (Table 7). In general, percent occurrence of ring nematode was almost always greater in healthy Rhododendron plots than in dieback plots. For the other two nematode species, no trends were observed. Meloidogyne sp. (root knot) did show a slight trend of having greater frequencies in healthy plots. Especially at Albert Mountain where many plants in dieback plots were dead, lack of available roots for the nematodes to reproduce in can have an inverse correlation with greater levels of dieback within plots. Data for the Albert Mountain plots showed that total number of ring nematodes averaged across three years and treatments was 6959 compared to Laurel Falls where the average was 6399 per 473 cc soil. Rotylenchulus reniformis (reniform), which had much lower numbers at Albert Mountain in the dieback plots, had a mean average of 2.8 and occurred in healthy plots at an average of 0.7. When the nematode populations were compared between Albert Mountain and Laurel Falls, significant levels occurred for root knot, Pratylenchus sp. (lesion), and Helicotylenchus sp. (spiral). Root knot had a mean average of 105.4 at Albert Mountain and 4.4 at Laurel Falls. Lesion averaged 16.5 at Albert Mountain and 0.4 at Laurel Falls, and spiral averaged 83.4 and 19.5 for Laurel Falls and Albert Mountain, respectively. When nematode levels were compared by year, Heterodera sp. (Cyst) had a significantly higher mean average (6.3) in 2008 but none were found in 2006 across locations. Also, we found reniform nematodes at significantly greater levels in 2006 (3.5) than in 2008 (0). These population mean totals would be considered below levels necessary to cause above ground symptoms. Nematode levels were significantly greater at Albert Mountain than at Laurel Falls for root knot, sting, and stunt nematodes (Table 8). Ring nematodes were numerically greater at Albert Mountain than at Laurel Falls. However, Laurel Falls had significantly greater population levels of spiral, lance, and cyst nematodes. Sting had low numbers at Albert Mountain where dieback plots showed severe Rhododendron losses. Two- and three-way interactions were noted for lesion and sheath (Yr * Loc.), ring (Yr. * Loc..; Yr. * Loc* Cond.), and reniform (Loc.* Cond.; Yr. Cond.; Yr.*Loc.*Cond.) nematodes, but these data did not show any meaningful interactions (data not shown). The sandy loam soil types at this high-elevation site are ideal for survival and reproduction of sting nematodes, but this soil type was limited to the upper ridge top in one large area. Presence of even small numbers of nematodes from this genus can cause major root damage making the host plants susceptible to microbial infections and to poor environmental conditions. When Southeastern Naturalist 15 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 Table 7. Percent frequenciesA (and total numbers) of nematode species present in dieback and healthy Rhododendron maximum (Rosebay Rhododendron) plots at two locations in the southern Appalachian Mountains. Ten soil core samples were random collected from five plots of R. maximum dieback and healthy plots from each location. Mel = Meloidogyne sp. (root knot), Pra = Pratylenchus sp. (lesion), Hel = Helicotylenchus sp. (spiral), P. m. = Paratrichodorus minor (stubby root), X. a. = Xiphinema americanum (dagger), C. x. = Criconemella xenoplax (ring), Hpo = Hoplolaimus sp. (lance), Het = Heterodera sp. (cyst), Tyl = Tylenchorhynchus sp. (stunt), R. r. = Rotylenchulus reniformis (reniform), and Bel = Belonolaimus sp. (sting). Nematode species percent frequencies (total numbers) Location Mel Pra Hel P. m. X. a. C. x. Hop Het Tyl R. r. Bel 2006 Albert Mtn. Dieback 5.3 (874) 0 (0) 3.1 (505) <`1.0 (96) 0 (0) 11.1 (1806) <1.0 (8) 0 (0) 0 (0) 0 (0) 0 (40) Healthy 22.4 (3637) 0 (0) <1.0 (55) < 1.0 (56) 0 (0) 56.1 (9096) 0 (0) 0 (0) 0 (0) <1.0 (40) <1.0 (40) Laurel Falls Dieback <1.0 (96) 0 (0) 18.0 (1892) 2.1 (223) 0 (0) 26.7 (2802) <1.0 (72) 0 (0) 0 (0) 2.0 (206) 0 (0) Healthy <1.0 (56) 0 (0) 19.2 (2024) 1.8 (197) 0 (0) 26.5 (2783) 1.4 (151) 0 (0) 0 (0) <1.0 (8) 0 (0) 2007 Albert Mtn. Dieback 1.2 (79) 3.5 (223) 7.8 (505) <`1.0 (32) 0 (0) 20.8 (1343) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Healthy 0 (0) 3.2 (212) 2.5 (166) < 1.0 (8) 0 (0) 60.2 (3887) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Laurel Falls Dieback 0 (0) 0 (0) 10.8 (1066) 0 (0) 0 (0) 21.0 (2067) 3.3 (326) 4.7 (466) 0 (0) 0 (0) 0 (0) Healthy 2.7 (267) 0 (0) 26.7 (2639) 0 (0) 0 (0) 27.3 (2688) 1.9 (191) 1.5 (150) 0 (0) 0 (0) 0 (0) 2008 Albert Mtn. Dieback 9.2 (1246) 4.1 (560) 2.8 (386) <1.0 (32) <1.0 (20) 15.1 (2057) 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) Healthy 19.7 (2670) 5.6 (758) 4.5 (758) <1.0 (8) 0 (0) 37.9 (2688) 0 (0) 0 (0) <1.0 (40) 0 (0) 0 (0) Laurel Falls Dieback 1.2 (158) <1.0 (32) 7.0 (908) <1.0 (95) <1.0 (33) 20.9 (2649) 2.5 (324) 3.2 (419) 0 (0) 0 (0) 0 (0) Healthy <1.0 (48) 0 (0) 12.3 (1596) 2.3 (293) 0 (0) 47.1 (6108) 1.8 (238) <1.0 (111) 0 (0) 0 (0) 0 (0) APercent frequencies of nematodes are based on total number of nematodes (total number)/473 cc soil across dieback and healthy (five replicates each) by year and location; percentage occurrence was calculated as the total number of hits per species, out of the total number of occurrences (hits) for all species; note: more than one hit was possible for each species computed based on multiple occurrences within a given plot. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 16 dieback and healthy plots were compared across pooled locations, root knot and ring populations were significantly greater in dieback than healthy plots (Table 8). Also, Belonolaimus sp. (sting) was identified from the two sandy loam soil types of Albert Mountain. This location had severe dieback compared to Laurel Falls, and even though the nematode population levels were low, thresholds for sting nematodes are usually at just one for agronomic crops. Their damage to crop and plant roots can be severe and they are suspected to be a major contributor to dieback at Albert Mountain, but further testing is needed to verify this hypothesis. A large amount of literature is available showing possible associations of nematode with woody hosts in forest ecosystems (Dreistadt et al. 1994, Ruehle 1973). However, only stunt nematode was specifically reported as a host on Rhododendron (Dreistadt et al. 1994). Almost all species of nematodes identified during the current study were reported in forest ecosystems except stubby root and sheath. Ruehle (1973) reported that certain nematode species predisposed the host roots to fungal or bacterial invasion resulting in decline of the Rhododendron over time. This hypothesis may be the most plausible explanation in the Rhododendron dieback problem in the southern Appalachian Mountains. Bloomberg (1968) reported Table 8. Comparison of mean number for each nematode speciesA compared between locations, years, and dieback and healthy R. maximum (Rosebay Rhododendron) at one location within Great Smoky Mountains and one in Nantahala National Forest. Mel = meloidogyne (root knot), Hel = Helicotylenchus (spiral) , P. m. = Paratrichodorus minor (stubby root), X. a. = iphinema americanum (dagger), Hop = Hoplolaimus (lance), Het = Heterodera (cyst), Tyl = Tylenchorhynchus (stunt), Bel = Belonolaimus (sting), Cysts = Cysts spp. (other). Nematode species (balanced mean data)B Mel Hel P. m. X. a. Hop Het Tyl Bel Cysts Plot location Albert Mountain 105.4 AC 19.5 B 3.3 B 0.1 A 0.1 B 0 B 0.6 A 0.5 A 2.1 A Laurel Falls 4.4 B 83.4 A 9.9 A 0.2 A 9.9 A 6.3 A 0.2 B 0 B 3.3 A LSD (62.4) (28.3) (8.2) (0.4) (12.1) (6.1) (1.3) (0.9) (2.1) Treatments/across locations Dieback 80.0 AD 51.3 A 7.7 A 0.1 A 4.9 A 1.4 A 0.5 A 0.5 A 3.3 A Healthy 29.7 A 51.6 A 5.5 A 0.2 A 5.2 A 5.3 A 0.3 A 0 A 2.3 A LSD (60.8) (57.6) (7.8) (0.4) (3.1) (7.3) (1.4) (0.9) (1.1) Years/across locations 2006 58.3 A 56.1 A 7.1 A 0 A 3.0 A 0 B 0.1 A 0.5 A 3.0 A 2008 51.4 A 46.9 A 6.2 A 0.3 A 7.0 A 6.6 A 0.7 A 0 B 2.6 A LSD (18.8) (42.6) (4.1) (0.3) (6.6) (6.0) (1.2) (1.9) (2.0) AAt the two locations each replicate plot (dieback and healthy) had 10 randomly collected soil samples during each year. BLesion and sheath nematodes were not included in the table since they had two-way interactions (Year*Loc), and reniform and ring had two- and three-way interactions ([Year*Loc] and [Year*Loc*Cond]). CMeans (based on 473 cc soil) within columns followed by the same letter are significantly different (P ≤ 0.05) according to the LSD. DTotal cysts were counted and then placed onto a Baermann funnel (Baermann funnel method) for five days; hatched larve were counted and identified as Heterodera spp. Southeastern Naturalist 17 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 that Xiphinema bakerii (Dagger) was consistently associated with brown rot of Pseudotsuga menzensii (Mirb.) Franco (Douglas Fir) roots, and that the pathogen Cylindrocarpon radicicola Wr. (= N. radicicola) forms a disease complex with this nematode species resulting in increased damage to host plants. Furthermore, Ruehle (1973) reported that C. xenoplax forms a disease complex with Fusarium solani (Mart.) Sacc. and Pythium irregulare Buis. It is possible that root knot and/ or ring nematodes formed an association with N. radicicola at both dieback sites causing increased decline of Rhododendron. In the current baseline study, N. radicicola was commonly isolated from the Rhododendron root tissues from the Rhododendron plants in the dieback and healthy plots. In this situation, the association with nematode species would not have been with X. bakerii or Xiphinema spp. due to low numbers found throughout the study. The interaction of the fungal pathogen with other nematode species such as ring nematodes should be investigated to confirm that Rhododendron root health may decline at higher rates due to changes in environmental conditions (e.g., high temperatures) in the forest ecosystem. Higher temperatures and impacts of reduced tree growth have previously been discussed (Ruehle 1973). Environmental changes to the forest ecosystem may be a key contributor to Rhododendron dieback but have yet to be verified. The impact of nematodes on mycorrhizal associations was discussed by Ruehle (1962). Ectotmycorrhizal fungi develop fungal mantles or “Hartig Nets” around the secondary and feeder roots of trees that form a physical barrier which can limit invasion by nematodes species and can reduce tissue damage and suppress the invasion of pathogenic fungi in many cases. The majority of forest trees form endomycorrhizae with fungi. No mantle is formed in this association, resulting in less protection from damage to roots by nematode species (Ruehle 1973). Root knot nematodes have been found forming galls on roots of Tulip Poplar, and endomycorrhizae could not be found in the root cells of the galled tissues (D. Marx, USDA/ARS-Athens, GA, unpubl.). In another study, Tulip Poplar cortical cells of feeder roots were being damaged by lesion nematodes, and normal endomycorrhzae associates failed to colonize the tissues or prevent lesions from forming (J. Ruehle, USDA/ARS-Athens, GA, unpubl.). Since fungi form Ericaceous mycorrhizal associations with Rhododendron spp., no mantles are formed and loss of endomycorrhizal associations in dieback plots may have occurred from the nematode pressure. Other reports of nematode species attacking forest trees and woody plants included Pratylenchus macrostylus Wu on Picea and Abies spp. of southern forests (Hartman and Eisenback 1991). In Slovakia, 29 parasitic nematode species were reported from 20 randomly sampled forest tree nurseries (Stollarova 1999). In the current study, stubby root and ring nematodes were shown to have the highest levels in dieback plots, but no clear trends occurred associating them with Rhododendron dieback. Therefore, microplot or controlled greenhouse studies must be conducted to determine the importance of these two nematode species with associated fungi, such as I. radicicola, previously reported to be important in the disease complex. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 18 When site and plant data were compared between locations, tree diameters were significantly greater in dieback plots at Albert Mountain than at Laurel Falls (Table 9). Furthermore, dieback plots across pooled locations had significantly greater tree diameters, with a mean average of 7.9 cm compared to 6.1 cm in healthy plots. Also, dieback ratings directly corresponded to greater diameters, but no differences in clonal unit numbers occurred between dieback and healthy plots. A preliminary study conducted in 1993 by Park Service staff in GRSM at Cerulean Knob showed that dbh of 30 Rhododendrons were not significantly correlated with age based on growth rings (G. Taylor, unpubl. data). These data may indicate a slower growth response in dieback areas occurring prior to visible symptoms due to increased environmental stresses with associated biotic organisms. Comparison of data between locations indicated that those on Albert Mountain occurred on a significantly different aspect—south-facing—compared to those on Laurel Falls, which were generally northern (Table 10). Previous research has shown that at Albert Mountain rainfall amounts were at least 38.1 cm less than over the ridge top area. With several years of drought and an unusual occurrence of sandy soils at this location, these attributes could have contributed to the increased stress to Rhododendron already under nematode and root pathogen pressure (Swift et al. 1988). Slope aspect modifies microclimate and influences ecological processes and spatial distribution of species across forest landscapes, but the impact of slope aspect on community responses to disturbance is poorly understood. In a boreal ecosystem, plant and bryophyte species changed more on south-facing slopes following clearcutting of Picea abies (L.) H. Karst (Norway Spruce) (Åström et al. 2006). Previous studies have shown that aspect, through its impact on microclimate, influences the spatial distribution of vascular plants (Bale et al. 1998, Cantlon 1953). Mean slope was significantly greater at dieback sites (22.8%) than at healthy sites (17.5%) across locations. When disease data were compared by location (dieback and healthy plots), average ratings were significantly greater for Albert Mountain (6.3) than Laurel Falls (5.2). Whereas numbers of Rhododendron clones Table 9. Comparison of R. maximum (Rosebay Rhododendron) diameters, dieback ratings, and numbers of clonal units at two locations and between dieback and healthy plots. Diameter (cm) Dieback R value rating Number of tree clones LocationB Albert Mountain 7.9 AA 4.8 A 2.8 A Laurel Falls 6.4 B 3.7 B 3.2 A LSD (1.42) (0.71) (0.80) Treatment Dieback 7.9 A 5.8 A 3.0 A Healthy 6.1 B 2.6 B 3.0 A LSD (1.4) (0.57) (0.89) AMean numbers followed by a different letter are significantly different (P ≤ 0.05) according to LSD. BAlbert Mountain is located in Nantahala National Forest, and Laurel Falls is in the Great Smoky Mountains National Park. Southeastern Naturalist 19 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 were similar at both locations, plant diameter and disease ratings were either numerically or significantly greater at Albert Mountain than Laurel Falls. The plant and disease data compared by years, across location, dieback versus healthy plots, showed no differences. Mean elevation data between the two locations was significantly different, with Albert Mountain at 1543.5 m and Laurel Falls at 946.4 m. There was significantly greater dieback at Albert Mountain than at Laurel Falls making it possible that elevation differences might impact dieback occurrences. This result might occur because high-elevation sites are not normally as suitable for Rhododendron growth as in riparian zones. It is more likely that elevation is an additional parameter that must be considered with other biotic and abiotic stresses to the Rhododendron, but further site comparisons are required. Routine soil-test procedures were conducted to elucidate any nutrient deficiency or soil pH effects that could contribute to the decline. Even though ranges of some of the soil properties varied widely (for example, exchangeable Ca2+ ranged from 11.2 to 218 kg/ha, exchangeable Mg2+ from 11.5 to 60 kg/ha, and soil pH from 3.81 to 5.25), there were no significant trends observed to suggest any of the measured soil factors contributed to the decline. Table 10. Comparison of R. maximum (Rosebay Rhododendron) site parameter mean dataA at two locations and between dieback and healthy plots. LocationB/treatment Aspect Soil depth (cm) Slope (%) Albert Mountain All treatments 1.8 B 111.8 A 26.7 A Laurel Falls All treatments 5.1 A 101.0 B 13.6 B LSD -0.71 -8.07 -5.69 Both locations combined Dieback 3.8 B 106.7 A 22.8 A Healthy 3.7 A 106.1 A 17.5 B LSD -0.47 -13.87 -2.93 Albert Mountain Dieback 2.0 A 114.7 A 18.4 A Healthy 1.6 A 108.9 A 8.8 B LSD -0.78 -16.12 -5 Laurel Falls Dieback 5.5 A 97.5 A 27.3 A Healthy 5.7 A 104.5 A 26.2 A LSD -2.45 -15.99 -7.47 AMean numbers followed by a different number across columns are significantly different (P ≤ 0.05) according to LSD. BAlbert Mountain is located in Nantahala National Forest, and Laurel Falls is in the Great Smoky Mountains National Park. CAspect numbers include 1 = north, 2 = northwest, 3 = northeast, 4 = south, 5 = southwest, 6 = southeast, 7 = east, and 8 = west; soil depth was determined by replicate readings taken with each subplot per plot; slope is in percent and was determined within each subplot per plot. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 20 Multiple regression analysis was done using all the data evaluated in this study. The relationship of disease rating (R_Calc) by select edaphic factors was evaluated for Albert Mountain and Laurel Falls (Tables 11, 12). Equations accounting for the greatest variations in dieback ratings were selected on basis of R2 of the regression equation. For Albert Mountain, factors affecting disease ratings included tree diameter (with higher ratings always associated with increased tree diameter), and nutrient levels of Na, P, and % Ca2+ of the CEC. Other factors important to disease rating were site factors such as aspect, soil depth and slope, and ring, root knot, and reniform nematode levels. Since the Albert Mountain site is on a ridge top and plots were scattered on different aspects, those data were evaluated. However, the R2 values were greatest for aspect most likely due to plot layout at this location. Population levels of ring and root knot nematode species were high at both locations and would be expected to cause sufficient root damage to Rhododendrons. It is uncertain if reniform nematode levels at both locations could impact plants. Overall, the three nematode species are believed to form a disease complex with microbes discussed previously. At Laurel Falls, factors impacting disease rating were exchangeable Ca2+, % Mg2+ saturation of the CEC, and estimated total CEC. In addition, dagger, lance, and reniform nematodes were found to impact the disease ratings at Laurel Falls. Over the three years of sampling, these three nematode species never had high population levels, and their correlation with increased dieback ratings is doubtful. In conclusion, presence of select fungi in association with several species of nematodes are believed to form a disease complex resulting in decline of Rhododendron trees, especially where they are growing outside their normal range such as on Albert Mountain. Furthermore, additional studies are necessary to confirm the role of important nematode species in association with fungi such as N. radicicola with a putative disease complex reported previously in the literature. Greenhouse and additional field studies will be necessary to evaluate the disease complex hypothesis. Furthermore, these baseline data of belowground fungal communities are Table 11. Stepwise regression equations for the variable R_Calc as affected by soil nutrients and nematode densities from the Laurel Falls location. Equation Step R_Calc with nutrientsA R_Calc with nematodesB 1 y = 1.82755 - 6.76746M y = 3.62012 + 13.78594D R2 = 0.2116, P = 0.0730 R2 = 0.2368, P = 0.0559 2 y = 1.82755 - 6.76746M + 6.26613P y = 3.62012 + 13.78594D + 0.09179R R2 = 0.4243, P = 0.0472 R2 = 0.3840, P = 0.1015 3 y = 3.62012 + 13.78594D + 0.09179R - 0.03728L R2 = 0.4859, P = 0.1490 A M = MQCA or Magnesium milliequivalent per 100 grams soil, P = PCMG or percent of magnesium of estimation of total cation exchange capacity (ECEC). B D = dagger nematode, L = lance nematode, R = reniform. Southeastern Naturalist 21 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 Table 11. Stepwise regression equations for the variable R_calc as affected by above ground variables, soil nutrients, aspect, and nematode densities from the Albert Mountain location. Equation Step R_Calc with above groundA R_Calc with nutrientsB R_Calc with aspectC R_Calc with nematodesD 1 y = 4.35676 - 1.25021C y = 2.44866 + 0.03567Na y = -2.01516 + 0.21852S y = 6.82024 - 0.00627R R2 = 0.2999, P = 0.0281 R2 = 0.2208, P = 0.0663 R2 = 0.8131, P = 0.0022 R2 = 0.4030, P = 0.0082 2 y = 4.35676 - 1.25021C + 0.44368D y = 2.44866 + 0.03567Na - 2.11009P y = -2.01516 + 0.21852S + 2.29672A y = 6.82024 - 0.00627R - 0.00653RK R2 = 0.4269, P = 0.1135 R2 = 0.3903, P = 0.0797 R2 = 0.9140; P = 0.0600 R2 = 0.5387, P = 0.0724 3 y = -2.01516 + 0.21852S + 2.29672A y = 6.82024 - 0.00627R - + 0.04277D 0.00653RK - 2.42501Re R2 = 0.9637, P = 0.0791 R2 = 0.6569, P = 0.0647 A D = diameter in cm, C = number of clones. B N = Na for sodium , P = PCCA for percent of calcium of estimation of total cation exchange capacity (ECEC). CSite factors including S = slope, A = aspect, D = depth. DNematode species were R = ring nematode, RK = root knot nematode, Re = reniform. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 22 the first ever recorded for Rhododendron habitats within the southern United States and within the Appalachian Mountain range. Acknowledgments Appreciation is extended to the Great Smoky Mountains Conservation Association for a Carlos C. Campbell Memorial Fellowship in 2007. In addition, financial support of the project was provided by the American Rhododendron Society of that year. A thank you is extended to Highlands Biological Station for financial support as Grants-In-Aid during 2007/2008, and David Pratt, Facilities Manager, University of Tennessee Field Biology Station, for housing and laboratory use during the project. Finally appreciation is extended to Mississippi State University (MAFES publication number 12251) for use of laboratory facilities and random supplies not covered by grants. Literature Cited Åström, M., M. Dynesius, K. Hylander, and C. Nilsson. 2006. Slope aspect modifies community response to clear-cutting in boreal forests. Ecology 88:749–758. Baird, R., P. Wadl, T. Rinehart, H. Abbas, T. Shier, and R. Trigiano. 2010. SSR marker for genetic comparison of Macrophomina phaseolina isolates from different states and hosts throughout the continental United States. Mycopathologia 170:169–180. Baird, R., A. Wood-Jones, J. Varco, W. Starrett, G. Taylor, and K. Johnson. 2013. Rhododendron decline in Great Smoky Mountains and surrounding areas: Evaluation of select parameters associated with decline. Southeastern Naturalist 12(4):703–722. Bale, C.L., J.B. Williams, and J.L. Charley. 1998. The impact of aspect on forest structure and floristics in some eastern Australian sites. Forest Ecology and Management 110:363–377. Baker, K.R. 1978. Determining nematode population responses to control agents. Pp. 114–127, In E.I. Zehr (Ed.). Methods for Evaluating Plant Fungicides, Nematicides, and Bactericides. American Phytopathological Society, St. Paul, MN. Barker, K.R. 1978. Determining nematode population responses to control agents. Pp. 114–127, In E.I. Zehr (Ed.). Methods for evaluating plant fungicides, nematicides, and bactericides. American Phytopathological Society, St. Paul, MN. Barnett, H.L., and B.B. Hunter. 1998. Illustrated Genera of Imperfect Fungi, 4th Edition. Macmillian Publishing C.ompany, New York, NY. 218 pp. Benson, D.M., and K.R. Barker. 1988. Diseases caused by nematodes. Pp. 34–35, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. Benson, D.M., and H.A.J. Hoitink. 1988. Phytophthora dieback. Pp. 12–15, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. Bernard, E.C. 2005. Nematodes—a one day sprint. University of Tennessee, Knoxville, TN Agricultural Experiment Station. Available online at http://eppserver.ag.utk.edu/ courses/EPP520/Nematodes_files/ frame.htm. Accessed 20 November 2005. Bloomberg, W.J. 1968. Corky root disease of Douglas Fir seedlings. Canadian Department of Forestry and Rural Development, Bi-monthly Resource Notes 24:8. Boettcher, S.E., and P.J. Kalisz. 1990. Single-tree influence on soil properties in the mountains of eastern Kentucky. Ecology 71:1365–1372. Borneman, J., and J.R. Hartin. 2000. PCR primers that amplify fungal rRNA genes from environmental samples. Applied Environmental Microbiology 66:4356–4360. Southeastern Naturalist 23 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 Brown, M.L. 2000. The Wild East: A Biography of the Great Smoky Mountains. University Press of Florida, Gainesville, FL. 457 pp. Cantlon, J.E. 1953. Vegetation and microclimates on north and south slopes of Cushetunk Mountain, New Jersey. Ecological Monographs 23:241–270. Cox, M.S. 2001. The Lancaster soil test method as an alternative to Mehlich 3 soil test method. Soil Science 166:484–489. Coyier, D.L., and M.K. Roane (Eds.). 1988. Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. 65 pp. Dighton, J., and A.F. Harrison. 1983. Phosphorus nutrition of Lodgepole Pine and Sitka Spruce stands as indicated by a root bioassay. Forestry 56:33–43. Dreidstadt, S.H., J.K. Clark, and M.L. Flint. 1994. Pests of landscape trees and shrubs. Pp. 65–92, In C. Laning (Ed.). An integrated pest management guide. Publ. 3359. University of California, Division of Natural Resources, Davis, CA. Ellis, M.B. 1971. Dematiaceous Hyphomycetes. Commonwealth Mycological Institute, KEW, Surrey, UK. 608 pp. Ferguson, A.J., and S.N. Jeffers. 1999. Detecting multiple species of Phytophthora in container mixes from ornamental crop nurseries. Plant Disease 83:1129–1136. Fink, S. 1999. Pathological and Regenerative Plant Anatomy. Gebrüder Borntraeger, Berlin, Germany. 1095 pp. Gardes, M., and T.D. Bruns. 1993. ITS primers with enhanced specificity for basidiomycetes- application to the identification of mycorrhizae and rusts. Molecular Evolution 2:113–118. Geiser D.M., M. del Mar Jimenez-Gasco, S. Kang, I. Makalowska, N. Veeraraghavan, T.J. Ward, N. Zhang, G.A. Kuldau, and K. O’Donnell. 2004. FUSARIUM-ID v. 1.0: A DNA sequence database for identifying Fusarium. European Journal Plant Pathology 110:473–479. Goodman, D.M., D.M. Durall, J.A. Trofymow, and S.M. Berch. 1996. Concise Descriptions of North American Ectomycorrhizae. Mycologue Publications, Sidney, BC, Canada. Folios 1–5. Grelet, G.A., D. Johnson, T. Vralstad, I.J. Alexander, and I.C. Anderson. 2010. New insights into the mycorrhizal Rhizoscyphus ericae aggregate: Spatial structure and co-colonization of ectomycorrhizal and ericoid roots. New Phytologist 188:210–222. Hartman, K.M., and J.D. Eisenback. 1991. Amended description of Pratylenchus macrostylus Wu, 1971 with SEM observations. Journal of Nematology 23:104–109. Hoitink, H.A.J., D.M. Benson, and A.F. Schmitthenner. 1988. Phytophthora root rot. Pp. 4–8, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. Jeffers, S.N., and H.S. Aldwinckle. 1987. Enhancing detection of Phytophthora cactorum in naturally infested soil. Phytopathology 77:1475–1482. Jeffers, S.N., and S.B. Martin. 1986. Comparison of two media selective for Phytophthora and Pythium species. Plant Disease 70:1038–1043. Jones, R.K. 1988. Botryosphaeria dieback. Pp. 10–11, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. Leyv, E.M., G. Zilberstaine, G. Elkind, M. Zeidan, E. Teverovsky, and I.S. Ben-Ze’ev. 2007. First report of Neonectria radicocola on Avocado in Israel. Phyoparasitica 35:2. Lickey, E., S.M. Tieken, K.W. Hughes, and R.H. Petersen. 2007. The mushroom TWIG: A marvelous mycological menagerie in the mountains. Southeastern Naturalist 6 (Special issue 1):73–82. Linderman, R.G. 1988. Phytophthora syringae blight. Pp. 15–17, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. Southeastern Naturalist R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 24 Madden, M., R. Welch, T. Jordan, P. Jackson, R. Seavey, and J. Seavey. 2004. Digital vegetation maps for the Great Smoky Mountains National Park. Final Report. The University of Georgia, Athens, GA. 44 pp. Mata, J.L. K. Hughes, and R.H. Petersen. 2007. An investigation of Omphalotaceae (Fungi: Euagarics) with emphasis on the genus Gymnopus. Syndowia 58:191–289. Mississippi State University (MSU). 2004. Soil testing for the farmer. Available online at from http://msucares.com/pubs/infosheets/is0346.htm. Accessed 3 November 2005. Mississippi State University (MSU). 2005. Plant disease and nematode diagnostic services. M-1230. Mississippi State Extension Services, Mississippi State, MS. Moore, L.W. 1988. Diseases caused by bacteria. Pp. 29–30, In D.L. Coyier and M.K. Roane (Eds.). Compendium of Rhododendron and Azalea Diseases. APS Press, St. Paul, MN. O’Brien, H.E., J.L.Parrent, J.A.Jackson, J-M. Moncalvo, and R. Vilgalys. 2005. Fungal community analysis by large-scale sequencing of environmental samples. Applied Environmental Microbiology 71:5544–5550. Ownley, B.H. 1993. Investigator’s Annual Report-Part II. National Park Services, US Department of the Interior. GRSM-N-0860083. Gatlinburg, TN. Ownley, B.H. 1994. Biotic and abiotic factors associated with rhododendron dieback in the Great Smoky Mountains National Park. National Park Services, US Department of the Interior. Sub-Agreement to Cooperative Agreement CA-5460-0-9001. Gatlinburg, TN. Pickering, J., R. Kays, A. Meier, S. Andrew, and R. Yatskievych. 2003. The Appalachians. Pp. 458–467, In C. Goettsch Mittermeir, G. Fonseca, J. Pilgram, T. Brooks, W. Konstant, P. Robles Gil, and S. De. (Eds.). Wilderness: Earth’s Last Places. Conservation International, Agrupacion Sierra Madre, Mexico City, Mexico. Pearson, V., and D.J. Read. 1973. The biology of mycorrhiza in the Ericaceae I. The isolation of the endophyte and synthesis of mycorrhizas in aseptic culture. New Phytology 72:371–379. Rossman, A.Y., G.J. Samuels, C.T. Rogerson, and R. Lowen. 1999. Genera of Bionectriaceae, Hypocreaceae, and Nectriaceae (Hypocreales, Ascomycetes). Studies in Mycology 42. CBS Press, Utrecht, Uppsalalaan, The Netherlands. 248 pp. Ruehle, J.L. 1962. Histological studies of pine roots infected with lance and pine cytoid nematodes. Phytopathology 52:68–71. Ruehle, J.L. 1973. Nematodes and forest trees-types of damage to tree roots. Annual Review Phytopathology 11:99–118. Sabanadzovic, S., S. Abou Ghanen–Sabanadzovic, and N. Valverde. 2010. A novel monpartite dsRNA virus from Rhododendron. Archives Virology 155:1859–1863. SAS Institute. 1999. SAS/STAT software changes and enhancements through release V8. Cary, NC Sinclair, W.A., H.H. Lyon, and W.T. Johnson. 1993. Diseases of Trees and Shrubs. Comstock Publishing Associates, Ithaca, NY. 575 pp. Stephenson, S.L. 1989. Distribution and ecology of myxomycetes in temperate forests II. Patterns of occurrence on bark surface of living trees, leaf litter, and dung. Mycologia 81:608–621. Stephenson, S.L., M. Schnittler, and C. Lado. 2004. Ecological characterization of tropical myxomycete assemblage-Maquipucuna Cloud Forest Reserve, Equador. Mycologia 96:488–497. Stollarova, I. 1999. The occurrence, distribution, and abundance of plant parasitic nematodes in forest and fruit nurseries of Slovakia. Nematologia Mediterranean 27:47–56. Sutton, S.C. 1980. Coelomycetes. Commonwealth Mycological Institute, Kew, Surrey, UK. 696 pp. Southeastern Naturalist 25 R. Baird, A. Wood-Jones, J. Varco, C. Watson, W. Starrett, G. Taylor, and K. Johnson 2014 Vol. 13, No. 1 Swift, L.W., Jr., G.B. Cunningham, and J.E. Douglass. 1988. Climatology and hydrology. Pp. 35–55, In W.T. Swank and D.A. Crossley, Jr. (Eds.). Forest Hydrology and Ecology at Coweeta. Ecological Studies, Vol. 66. Springer‑Verlag. New York, NY. Toussoun, T.A., and P.E. Nelson. 1976. Fusarium: A Pictorial Guide to the Identification of Fusarium Species According to the Taxonomic System of Snyder and Hansen. 2nd Edition. The Pennsylvania State University Press, University Park, PA. 43 pp. Tuite, J. 1969. Plant Pathological Methods: Fungi and Bacteria. Burgess Publishing Company, Minneapolis, MN. 239 pp. Van Dyke, F. 2003. Conservation Biology: Foundations, Concepts, Applications. McGraw Hill, Boston, MA. 413 pp. Villarroel, D.A., R.E. Baird, L.E. Trevathan, C.E. Watson, and M.L. Scruggs. 2004. Pod and seed mycoflora on transgenic and conventional soybean (Glycine max (L.) Merrill) cultivars in Mississippi. Mycopathologia 157:207–215. Walker, J.F., L. Aldrich-Wolfe, A. Riffel, H. Barbare, N.B. Simpson, J. Trowbridge, and A. Jumpponen. 2011. Diverse Helotiales associated with roots of three of Artic Ericaceae provide no evidence for host specificity. New Phytologist 191:515–527. Watanabe, T. 1994. Pictorial Atlas of Soil and Seed Fungi. CRC Press, Boca Raton, FL. 411 pp. White, T.J., T. Brun, and L.S. Taylor. 1990. Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. Pp. 315–322, In M.A. Innis, D.H. Gelfand, J.J. Sninsky, and T.J. White (Eds.). PCR Protocols: A Guide to Methods and Applications. Academic Press, New York, NY.