Coral Resistance to Disease

22

Coral Resistance to Disease Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

22.1 Theoretical Framework Understanding the dynamics of resistance is particularly important for understanding the impacts of disease and predicting evolutionary outcomes for diseases. Predictive epidemiological models include not only terms for transmission of infectious microorganisms, but also terms for host resistance. In susceptible-infected-resistant (SIR) epidemiological models, timing and degree of resistance can determine the spread rate and impact of disease (Anderson and May 1979, 1991). Resistance is defined as “the natural or acquired ability of an organism to maintain its immunity to or to resist the effects of an antagonistic agent, e.g., pathogenic microorganism, toxin, drug (Stedman 1995).” An organism that is immune to an infectious disease will not acquire it because it has a particular suite of complex structural and functional features. These features prevent the pathogenic microorganism from entering, surviving in, or multiplying within its body and causing disease by disrupting key cellular metabolic processes through the release of toxins or enzymes or by altering its structure (e.g., tissue damage through scarring), or causing cell death. Many factors can affect the condition of this system and the response to a pathogen that an individual host is capable of generating at a particular time. The interaction of host and pathogen, and how they are affected by changing environmental conditions, can affect the populations of both organisms (Garnett and Holmes 1996). Understanding the mechanisms of coral resistance to disease is of particular importance because in warming oceans, corals are demonstrably stressed by high summer temperatures. Stress in corals can be identified by an increased rate of bleaching (Hoegh-Guldberg 1999; Bruno et al. 2001; see other chapters in this Vol.), which may be linked to the appearance of some diseases (Kushmaro et al. 1997; Harvell et al. 2001; Porter et al. 2001), suggesting a role for compromised resistance. In some cases, bleaching itself is an infectious disease (Kushmaro et al. 1997; Ben-Haim et al. 1999; Ben-Haim and Rosenberg 2002). The rates of coral bleaching have increased in the last three decades and impacts of coral disease also appear to have increased (Santavy and Peters 1997; Hoegh-Guldberg 1999; Porter et al. 2001; Bruckner 2002; Ward and Lafferty 2004).

378

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

22.2 Known Mechanisms of Coral Resistance Although current rates and impacts of disease in corals are high (Aronson and Precht 1997; Richardson 1998; Richardson et al. 1998; Harvell et al. 1999; Weil et al. 2000, 2001; Porter et al. 2001), little is known about the resistance of coral to infectious disease. Bigger and Hildemann (1982) reviewed cellular defense systems of the Cnidaria, including pathogen defense, wound healing and inflammation, and response to foreign tissue. There is no previous work on coral resistance to pathogen infections, except for reviews of generalized coral responses to stress and injury (Peters 1984b; Hayes and Goreau 1998; Olano and Bigger 2000). Recent experiments and histological observations of scleractinian (Hexacorallia) and gorgonian (Octocorallia) corals provide insights into how resistant these organisms might be to pathogenic microorganisms. 22.2.1 Structure and Function of Coral Cells The anatomy and histology of corals have been described by Hyman (1940), Bayer (1974), Chapman (1974), Peters (1984a), Fautin and Mariscal (1991), and others. The basic structure in each group is the polyp, a hollow cylindrical blind-ended sac like a sea anemone, often connected to other polyps by gastrovascular tissue, forming a colony The polyp has a mouth, surrounded by a ring of hollow retractable tentacles, and connected to the gastric cavity by a pharynx. The internal gastric cavity is divided by partitions called mesenteries. The mesenteries connect to the pharynx; within the gastric cavity the free edges of the mesenteries form mesenterial filaments. Colony formation differs between the groups. For scleractinia, the bases of the polyp sacs are embedded in the aragonite exoskeleton produced by the calicoblastic epithelium of the polyps, which lines the skeleton everywhere. In the octocorals, the bases of the polyp sacs are embedded in a thick layer of the primitive connective tissue known as mesoglea. Scleroblasts, modified epithelial cells within the mesoglea, form calcium carbonate sclerites varying in morphology from thin, spindle-shaped to thick, polymorphic, with variable surface projections to support and protect the tissue from predators. The horny corals or gorgonians are further supported by a proteinaceous rod produced by the axis epithelium. Polyps are connected to one another by cell-lined tubes known as gastrovascular canals in the scleractinia and solenia in the octocorals. The polyps are attached to their supporting exoskeletons or axial rods by cells called desmocytes (Bayer 1974; Muscatine et al. 1997). In both groups, a simple columnar or pseudostratified columnar epithelium, the epidermis, covers the external surfaces of the polyps and interpolypal tissue or coenosarc (coenenchyme). This epithelium covers the layer of mesoglea. Internally, the gastric cavity and canals that connect the polyps are lined by a generally cuboidal epithelium, the gastrodermis, also covering the mesoglea. The

22. Coral Resistance to Disease

379

mesenteries and their filaments within the gastric cavity are lined on both surfaces by gastrodermis with mesoglea between. The external and internal epithelial layers consist of several types of cells which provide protection or enable the polyps to capture and digest food, and support dinoflagellate algal cells that have a symbiotic relationship with their host coral (mainly in tropical shallow-water species). These algal cells or zooxanthellae are phagocytosed into vacuoles within the gastrodermal cells, but are not digested. They undergo photosynthesis in light and exchange nutrients and waste molecules with the polyp cells. Epitheliomuscular cells or myonemes and the subepidermal nerve net provide the polyps with the ability to expand or contract their bodies and tentacles, detect changes in the environment, and communicate with other polyps in the colony. Although the nematocysts and spirocysts are important in capturing zooplankton prey and protecting the coral from predators, they probably do not provide much protection against pathogenic microorganisms (viruses, bacteria, fungi, protozoa). The surface epidermis of scleractinia contains unicellular secretory or gland cells and ciliated supporting cells. These cell types can be reduced in size or fewer in number in the gorgonian epidermis. The gastrodermis also contains supporting and gland cells. The mesoglea binding the two layers of epithelia together throughout the colony consists of a gelatinous substance, collagen fibers, and cells. Although generally referred to as mesogleal cells, they represent different cell populations. Some appear to be fibroblasts and secrete the matrix and collagen fibers; others, called amoebocytes, can be granular or agranular and function as phagocytes (Bigger 1984; Olano and Bigger 2000). Some of these cells have also been identified as pluripotential stem cells, capable of dividing and differentiating into various cell types as needed, such as cnidoblasts, scleroblasts, or germ cells. The latter two groups are capable of migrating through the mesoglea to distant locations when needed in the epithelia. 22.2.2 Innate Immune Response Like other invertebrates, corals possess innate or natural immunity, a nonspecific ability to react to many potentially pathogenic organisms that is not altered with subsequent exposure. Basic host defenses include mechanical or physical barriers (e.g., epidermis), the ability to move to shed or expel pathogens, secretion of chemicals (e.g., acid) or production of bioactive compounds (e.g., antimicrobial peptides), and phagocytic cells that can engulf and destroy microorganisms on contact (Cotran et al. 1999). The cellular response consists of fixed or circulating amoeboid phagocytes that ingest microscopic organisms and kill them by exposure to proteolytic enzymes and free oxygen radicals. These cells go by different names in different phyla, e.g., leukocytes (macrophages) in vertebrates, hemocytes in mollusks, coelomocytes in echinoderms. For larger tissue-invading organisms, the amoe-

380

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

bocytes can surround the foreign form to encapsulate or wall it off, or form a nodule, an aggregation of amoebocytes and bacteria or other pathogenic microorganisms; these structures can be accompanied by the deposition of a layer of melanin. The humoral response in innate immunity consists of secreted antimicrobial peptides, macrokines (similar to cytokines), and lectins (to agglutinate microorganisms to make them easier to phagocytose). Acquired or adaptive immunity, cell-mediated and humoral, involves the production of specific antibodies and T lymphocytes to eliminate the invading microorganisms through the operation of the major histocompatibility complex restriction that protects normal cells from attack (Clancy 1998). Adaptive immunity against pathogenic microorganisms has not been demonstrated in invertebrates. Corals are animals, but because of their sessile nature and symbiosis with carbon-fixing algae, they have many plant-like physiological qualities. Therefore, in mapping out components of coral resistance to disease, it is useful to consider both plant and animal models. Plant inducible responses to fungi include constitutive and inducible components (Levin 1976; Agrawal et al. 1999; Berenbaum and Zangerl 1999). The main components of pathogen resistance are inducible and were classified by Kombrink and Somssich (1995), depending on speed of response and localization. Immediate early responses involve recognition and signaling processes, followed by locally initiated mechanisms such as phenylproponoid pathways, peroxidases and intracellular pathogenesis proteins. Finally, broad-spectrum systemic responses begin, such as production of chitinase and 1,2 beta-glucanases. Plant inducible responses to pathogens appear to diverge from responses to herbivores in using a salicylic acid pathway (Thaler et al. 2002a). Invertebrate defenses against microbial infections are diverse, as noted above, including largely inducible components such as encapsulation via prophenoloxidase (PPO)-catalyzed melanization (Aspan and Soderhall 1995), direct production of antimicrobial peptides, and multistep processes such as opsonization and phagocytosis initiated by lectin recognition. What is common to both plants and animals is the inducibility of the dominant mechanisms, rendering detection and timing of resistance components in corals a high priority. Many microorganisms have, however, developed their own protection against one or more of these defenses, with the result that infections and disease are present in host populations (Clancy 1998). Alternatively, anything that adversely affects the integrity of the coral cells or their ability to produce defense compounds by induction of key processes can permit infection by microorganisms and initiation of disease. 22.2.3 Coral Immune System Several studies have provided insights into how corals resist infection. For the sedentary scleractinian corals, the mucociliary system of the epidermis plays an important role in contrast to gorgonians. Mucous secretory cells are usually

22. Coral Resistance to Disease

381

abundant in the epidermis and copious quantities of mucus are released (Bigger and Hildemann 1982); the composition and structure of the mucus varies among species (Meikle et al. 1988). The acidic mucopolysaccharides can trap or repel bacteria; in other cases bacteria might use the compounds in the mucus as substrates (Rublee et al. 1980; Paul et al. 1986). Santavy (1995) noted that scleractinian corals infected by black-band disease had higher surface microbial productivity than healthy or otherwise compromised corals. Apical cilia on the supporting cells wave constantly, producing water flows to sweep mucus and trapped particles (e.g., bacteria, sediment) off the surface of the colony to fall to the base of the colony or be disbursed by reef currents. The production of mucus and ciliary beating require expenditure of much cell energy. Peters (1984b) found that the epidermis at the base or sediment margin of massive corals lacked mucous secretory cells, perhaps due to the constant work involved in trying to keep sediment off the coral. In a laboratory study, constant exposure to sedimentation for 3 months caused a reduction in the number of mucous secretory cells and changes in the pH of the mucus (Peters and Pilson 1985). Tissue loss due to sedimentation has been shown to be preventable in the laboratory when antibiotics are present (Hodgson 1990). Bacterial diseases such as white plague and black-band disease typically start at tissue margins (Antonius 1985; Richardson et al. 1998) where this defense could be weakened or nonexistent. Gorgonians, however, generally have fewer mucous secretory cells, although this depends on the species. Morphology of the colony, including vertical cylindrical growth to enable the polyps to extend into currents for food capture and maximum light exposure, also reduces the need for mucus. Cilia are present on cells of the epidermis, cnidoglandular tract of the mesenterial filaments, and pharynx to produce currents within the polyp to remove wastes. Phagocytosis is the dominant mechanism of defense in invertebrates. In Cnidaria, phagocytosis is accomplished by amoebocytes, motile phagocytic cells that take part in wound healing and tissue reorganization (Chapman 1974; Mattson 1976; Bigger and Hildemann 1982), as well as cells of the gastrodermis and epidermis when the host is traumatized (Olano and Bigger 2000). The amoebocytes can be agranular or contain numerous neutral or acidophilic granules under the light microscope. The density of the cells and their appearance varies between taxa as well as within colonies (Figs. 22.1, 22.2). Amoebocytes in the scleractinia are few and scattered within the mesoglea; they are best viewed in tissue sections of the fleshy species with larger polyps and thicker mesoglea. It is difficult to detect them in areas of thin mesoglea. In the mesoglea, they appear to be round to spindle-shaped, sometimes surrounded by a lacuna or space. The acidophilic granules have been considered to be lysosomes or peroxisomes (Olano and Bigger 2000). In the Gorgonia, amoebocytes occur in dense clusters throughout the thicker mesoglea. They can form a layer beneath the epidermis or be present between epidermal cells or on the surface of the epidermis, perhaps a first line of defense against bacteria through phagocytosis and within-cell destruction by enzymes.

382

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

n Fig. 22.1. Light microscopic view using oil immersion to show cells present in the mesoglea of a brain coral, Diploria strigosa. From left to right, the cells appear to be an agranular amoebocyte, a fibroblast, a stem cell, and a granular amoebocyte surrounded by a space

n Fig. 22.2. Light microscopic view using oil immersion to show cells present in the mesoglea of a sea fan, Gorgonia ventalina. In the center is a fibroblast, surrounded by acidophilic granular amoebocytes, much more numerous and larger than their scleractinian counterparts

The inflammatory process in which these cells participate is less well understood in invertebrates than vertebrates (Sparks 1972). Infiltration of phagocytic cells (macrophages) is one of the characteristics of inflammation. The roles of the different kinds of cnidarian amoebocytes have been postulated to include production of collagen fibers within the mesoglea (like fibroblasts); stem cells (sometimes referred to as interstitial cells in the literature) to differentiate into scleroblasts, germ cells, or other cell types; or assisting in wound

22. Coral Resistance to Disease

383

repair as phagocytes (increasing in numbers at wound sites as a result of migration), or differentiating into epidermal cells. In the Anthozoa, studies on phagocytosis and wound healing have been limited to sea anemones and gorgonians. Under normal conditions, the connective tissue of anemones contains a homogenous population of amoebocytes, but following wounding, cell density increased significantly in a circular pattern around the region of damage (Patterson and Landolt 1979). The amoebocytes had secondary lysosomes and were observed to behave as phagocytes, cleaning up damaged cells. Within the repair zone in the mesoglea, swelling of the mesoglea was found, along with diapedesis of phagocytes through mesoglea and epidermal cells to discharge debris at the surface, like that reported for mollusks. Phagocytes derived from amoebocytes infiltrated the mesoglea by migration from other sites (mitotic activity was not observed in these cells). The atypical cells found in the zone appeared also to be morphologically suited for the production and secretion of unknown substances. Finally, cells infiltrated the lesion from the surrounding epithelium. The authors noted that this was more than a simple phagocyte response and that a distinct series of cellular events followed this injury. They concluded that the anemone has a “functional inflammatory response that predates the origin of a circulatory system or specialized organs.” The inflammatory response in the gorgonian Plexaurella fusifera is also caused by amoebocyte accumulation at the wound site, an effect of cells migrating from adjacent uninjured tissue (Meszaros and Bigger 1999). The migration of amoebocytes into a wound region to isolate the damaged region, prevent secondary infection, and initiate tissue repair by producing mesogleal fibers is further evidence of an organized reaction to injury and infection (Meszaros and Bigger 1999). Despite numerous histological examinations of scleractinian corals affected by various lesions (wounding, tissue infiltration by algae, bleaching, and diseases such as black band disease and white band disease), inflammatory responses characterized by infiltration of numerous amoebocytes have not been detected. Both scleractinia and gorgonia are also capable of reacting to invading microorganisms by actively producing barriers to their penetration. For example, fungi that bore into the exoskeleton of scleractinians (Le Campion-Alsumard et al. 1995) induce activity by the calicoblasts, which lay down more skeleton. In histological preparations, the normally squamous calicoblastic epidermis becomes columnar with a more acidophilic staining cytoplasm adjacent to the fungal filaments. Layers of skeleton and organic material can be deposited to form a pearl. The axis epithelium and other cells of gorgonians can also be induced to begin more rapid production of gorgonin, with the deposition of melanin to wall off infiltrating fungi and algae (see below) and the formation of nodules (Morse et al. 1977). In addition to cell-mediated immune functions, corals produce antibacterial, antifungal, and predator-deterrent compounds (Jensen et al. 1996; Kim et al. 2000a, b). For example, the anemone Anthopleura elegantissima mucus con-

384

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

tains an enzyme that closely resembles lysozyme in its ability to lyse the bacterium Micrococcus lysodeikticus (Phillips 1963). Koh (1997) demonstrated that extracts from 100 coral species inhibited the growth of a marine cyanobacterium and extracts from eight of the species inhibited the growth of marine bacteria. Those eight species also had the fewest bacteria on their surfaces compared to corals lacking the antimicrobial compounds. Production of resistance compounds is also possible from associated surface bacteria. Twenty-nine percent of bacteria isolated from corals had antibacterial properties (Castillo et al. 2001). In other marine invertebrates, bacteria also appear as a source of antimicrobial compounds. Gil-Turnes et al. (1989) demonstrated that antifungal compounds that protect crustacean embryos from the fungal pathogen Lagenidium callinectes are produced by surface bacterial symbionts. The structural similarity between bryostatins of the bryozoan Bugula neritina and the bacterial symbiont Candidatus in Endobugula sertula suggests that the surface-associated bacteria produce the defensive compounds (Anthoni et al. 1990; Davidson and Haygood 1999). Among cnidarians, gorgonians display some of the most potent antimicrobial activities (Burkholder and Burkholder 1958; Burkholder 1973; Bigger and Hildemann 1982; Jensen et al. 1996; Kim et al. 2000a, b). Crassin acetate, found in the gorgonians Pseudoplexaura crassa and P. wagenaari and in the endosymbiotic zooxanthellae, has antimicrobial and antiprotozoan activity and deters parrotfish. The hydroquinones of Pseudopterogorgia rigida and P. acerosa have antiviral and antibacterial activity and deter predatory fish (Harvell et al. 1988). Immunoglobulin A was reported to be secreted by cnidarian mucous secretory cells (Tomasi and Grey 1972, cited in Hayes and Goreau 1998), but this has not been confirmed by others (see also Chap. 12, Kelman, this Vol. for antimicrobial compounds in corals.) The combination of cellular and humoral factors that make up the immune system varies from one individual to another; within the corals, it is clear that mucociliary activity, amoebocyte response, and production of antimicrobial compounds vary greatly among families, genera, and species. These genetically mediated differences might enable one group or one individual to have an advantage over others in resisting invasion by pathogens and reducing its susceptibility to disease. In addition, the age of the organism, its gender, reproductive state, and nutritional status can affect the immune system. For example, bleaching of tropical scleractinia or gorgonia for an extended period (weeks) removes a principal dietary resource, leading to atrophy and necrosis of the tissues (Lasker et al. 1984; Glynn et al. 1986). With loss of nutrients, mucus secretion, and ciliary beating, amoebocyte numbers are reduced, leaving polyps more susceptible to penetration by pathogenic microorganisms. Even if the polyps survive and recover their algal populations, reproduction and calcification can be inhibited for more than a year following the bleaching event, and other cellular processes might also be limited during this time (e.g., Szmant and Gassman 1990; Michalek-Wagner and Willis 2001; see also chapters in this Vol.). The line between reversible cellular changes and irreversible changes can

22. Coral Resistance to Disease

385

be difficult to distinguish, including those changes that affect the cells of the immune system in corals. Tissue loss during bleaching events might be due to host cell necrosis, or it might be due to pathogenic microorganisms that have easily evaded the weakened defenses. A priority is understanding what deficits occur during bleaching that might directly affect coral immunity and susceptibility to infection. Recent studies of aquatic organisms have sought to identify biomarkers, physiological, biochemical, or histological indicators, to show how well an organism’s immune system is functioning under different environmental conditions, or when exposed to pathogens. Hawkridge et al. (2000) identified several antioxidant enzymes mainly in intracellular granules, as well as in accumulation bodies of the zooxanthellae and in different types of cnidae, in the sea anemone Anemonia viridis and the scleractinian coral Goniopora stokesi. Downs et al. (2000) reported development of biomarkers in Montastrea faveolata to detect coral responses to thermal stress. These include molecular chaperones of temperature-sensitive pathways (heat shock proteins 60 and 70, chloroplast small heat shock protein), indicators of cell integrity (lipid peroxide, alpha beta crystalline, glutathione, and ubiquitin), and antioxidant enzymes indicative of oxidative stress (manganese superoxide dismutase, copper/zinc superoxide dismutase). These markers represent both zooxanthellae and coral stress proteins and respond to changes in temperature and light level. Downs et al. (2002) showed significant variation in these biomarkers for corals from different depths during a bleaching event, supporting the hypothesis that bleaching is driven by oxidative stress. Banin et al. (2000) detected toxin P as a virulence factor of Vibrio shiloi that inhibits photosynthesis of zooxanthellae. The presence of virulence factors that operate differentially on zooxanthellae and the coral host indicates that origins of resistance from both coral and zooxanthellae should be considered. In another experimental study of the basis of self-/nonself-recognition in the gorgonian Swiftia exserta, Salter-Cid and Bigger (1991) observed that histocompatibility reactions during tissue grafting met the minimal functional criteria of cytotoxicity, specificity, and altered secondary response (memory) that characterize an adaptive immune response. Autografts (host tissue applied to the same host) resulted in the fusion of the tissues. However, allografts (different donor tissue from the same species) resulted in rapid loss of tissue in the immediate contact area in 7–9 days. When another allograft was applied to the same host after a resting period, the same reaction occurred in only 3–4 days. Cell death was limited to the graft tissue interface, suggesting that this response was mediated by a contact or short-range cytotoxic molecule, rather than by a diffusible, long-range molecule (Salter-Cid and Bigger 1991). Additional studies are needed to confirm these observations. In summary, the immune system of corals shares similarities with other invertebrates, but is so poorly known that important differences might yet surface. The least understood components of coral immunity involve any possible collaboration between coral and algal cells and the role of the symbiosis in im-

386

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

n Fig. 22.3. Diagram of basic cellular changes occurring that adversely affect the host’s resistance as exposure to stressors increases

munity. With increasing exposure to environmental stressors outside the normal range to which an individual is accustomed (e.g., increases or decreases in salinity, oxygen, light; chemical contaminants), or to pathogenic microorganisms, the host’s immune system cells respond by undergoing detoxification or other metabolic reactions to try to reverse cellular changes and maintain the host organism’s homeostasis. These reactions can produce biomarkers, which can be measured to provide an indication of the functioning of the organism and its immune system. As the stressors continue to exert their effects on the cells, irreversible changes in the nucleus, organelles, and membranes can occur, signaling impairment of vital functions or systems (disease). Although the host immune response in invertebrates is simpler in concept than in vertebrates, we have much to learn about how the cells function and interact to provide resistance to diseases in corals (Fig. 22.3).

22.3 Gorgonians: the Sea Fan as a Model System In recent coral disease workshops (National Oceanic and Atmospheric Administration (NOAA) – Interagency Coral Disease and Health Consortium (CDHC), Charleston, SC, and World Bank, Akumal, Mexico), developing model systems for the study of coral resistance emerged as a research priority for future management and sustainability of reef habitats. A goal in our lab is to develop sea fans into such a model system to investigate chemical, cellular, and structural mechanisms of resistance. Critical priorities are to understand:

22. Coral Resistance to Disease

1. 2. 3. 4.

387

Cellular mechanisms of coral resistance; Chemical mechanisms of coral resistance; Relative contributions of zooxanthellae and corals to resistance; and Genetic variation within and among colonies in resistance.

The focus of this review will be to identify what we know about each of these critical areas and then suggest future directions in research. Aspergillosis is a disease of sea fan corals, first reported by Nagelkerken et al. (1996) and Smith et al. (1996). Because fungi in the genus Aspergillus are opportunistic pathogens in immune-compromised humans and other animals, the interaction between disease and resistance is of particular interest in this new outbreak in sea fans. We have shown that sea fans and other gorgonian coral species employ a battery of general antifungal and antibacterial compounds (i.e., secondary chemistry) for disease resistance (Jensen et al. 1996; Kim et al. 2000a, b), and have identified chitinase as a component of resistance extracts. In both plants and animals, systemic responses include hydrolytic enzymes such as chitinase (Tuzun and Bent 1999), a class of enzymes that hydrolyze chitin. Chitinases defend against fungal pathogens by destroying chitincontaining cell walls (Jolles and Muzzarelli 1999). Chitinolytic proteins are prominent, inducible components of antifungal resistance against Aspergillus fumigatus in guinea pigs (Overdijk et al.1996) and humans (Tjoelker et al.2000). Field and laboratory studies have shown variability among fans in host resistance and aggregation of diseased individuals. Dube et al. (2002) detected significant differences in mean and variance of antifungal activity (AFA) for sea fan populations at different locations in the Florida Keys. They also detected a correlation between disease pressure and variance in antifungal activity that is consistent with selection acting on antifungal activity. Jolles et al. (2002) mapped all fans within three replicate 10×10 m grids to investigate spatial distribution of infected fans. Using geostatistical analyses to separate aggregation of diseased from possible underlying aggregation of all fans, they detected significant aggregation of diseased fans. This aggregation could be caused by either secondary transmission among neighbors or variation in resistance. Because the degree of aggregation increases with increasing disease severity in this dataset, it seems more likely that aggregation is caused by factors affecting resistance. However, it is still not possible to rule out increased transmission in more aggregated locations as a cause of more severe disease. To understand the relationship between disease outbreak and resistance response requires an experimental approach. Because Aspergillus sydowii can be readily cultured, this patho-system allows development of challenge inoculation experiments. The protocol we have developed involves growing A. sydowii on PYG agar (0.2% peptone, 0.2% yeast extract, 0.5% glucose, 3.6% bactoagar, 0.005% tetracycline) into which sterile cotton wicks are embedded. The wicks can then be applied to sea fans (and other gorgonians) in the lab and field to test response to infection. For field experiments, we were cautious in applying pure isolates of A. sydowii isolated from those same reefs. Using these inocula-

388

Kerri M. Mullen, Esther C. Peters, C. Drew Harvell

tion protocols, we inoculated clonally replicated arrays of sea fans and showed that the level of AFA increased in inoculated fans and was higher in some clones (all the pieces from the same fan) than others (Harvell et al., unpubl.). This is the first experimental evidence for inducible AFA and for variation in levels of resistance among sea fans. Because corals are sessile-like plants, there is considerable insight to be gained from plant studies about the importance of genetic neighborhoods and resistance structure of hosts under disease pressure. Studies on the anther smut disease Usatilago violacea and the dioecious perennial Silene alba, have shown the importance of fungal pathogen and host genetic neighborhoods and frequency-dependent selection (Antonovics and Thrall 1994; Thrall and Burdon 2003). Studies of disease spread in experimental populations of S. alba, where transmission rates were manipulated by varying genetically based host resistance, have confirmed the importance of frequency-dependent selection in this system (Thrall and Jarosz 1994). Host genetic structure was manipulated by establishing relatively resistant and susceptible host families. The progeny of susceptible families had higher infection levels than those from resistant families, and both frequency and density of hosts affected disease spread. More experimental field studies of coral resistance are needed to fill in this type of spatial detail for corals. In our studies of resistance to fungal disease in gorgonians, we have identified several components 22.3.1 Generalized Antifungal Activity Minimum inhibitory concentration (MIC) assays showed that of the 20 common gorgonian species in the Florida Keys, extracts from 15 species had MICs