In 1944, a wartime supply ship anchored off Campbell Island. A meteorologist stepped ashore. He walked inland. He saw something growing among the Dracophyllum. Branched. Coral-like. Pale yellow. He collected it. The specimen dried. It sat in a herbarium for thirty years. In 1974, a mycologist opened the packet. She recognised it as new. She named it Ramaria subantarctica. The subantarctic coral. No one has collected it many times since. The islands are remote. The weather is terrible. The fungus fruits rarely. Each specimen is a event. Each collection a journey. The Subantarctic Coral is mycorrhizal. It forms a mutualism with the roots of subantarctic shrubs. Dracophyllum longifolium. Dracophyllum scoparium. The stunted Coprosma species that hug the ground. The fungus provides water and nutrients. The shrubs provide sugars. The exchange happens underground. The coral is the fruiting body. It emerges in summer. January. February. The days are long. The temperature is cold. Five degrees. Ten at most. The coral branches repeatedly. Each branch divides into two. Then two again. The tips are blunt. Pale yellow. Sometimes white. The base is thick. Fleshy. The whole structure reaches eight centimetres. Small for a Ramaria. Large for a subantarctic fungus. Habitat is restricted to the subantarctic islands of New Zealand. Campbell Island. The Auckland Islands. The Snares. The Antipodes Islands. The fungus grows on peaty soil. On slopes with good drainage. Not in swamps. Not on exposed ridges. It needs the shelter of the shrub canopy. The canopy is low. One metre. Two metres at most. The coral pokes through the leaves. It is visible. Barely. Most specimens are hidden. You must part the Dracophyllum branches. You must kneel. You must look closely. The fungus rewards patience. The branches are delicate. The colour is subtle. The form is precise. Range is confined to the subantarctic zone. The fungus does not occur on mainland New Zealand. Not in Fiordland. Not on Stewart Island. The climate is wrong. Warmer. Less windy. The soil chemistry is wrong. Less organic matter. More mineral content. The host shrubs are absent. Dracophyllum on the mainland grows taller. Its roots are different. The mycorrhizal partnership does not form. The coral stays where it evolved. The islands are refuges. The fungus has been there for millennia. Since the last glacial maximum. Since sea levels rose. Since the islands were cut off. The separation is complete. The fungus is trapped. It does not seem to mind. Threats are biological and climatic. Introduced mice occur on Campbell Island. They eat fungal tissue. They may eat the coral. No one has observed it directly. The mice are small. The coral is small. The match is plausible. Rats were eradicated from Campbell Island in 2001. Mice remain. The second threat is climate change. The subantarctic zone is warming. The peat soils are drying. Drier summers mean less fruiting. The mycelium survives. The coral waits. The wait may become permanent. The third threat is stochastic. A single shipwreck. A single fire. A single landslide. The islands are small. The populations are small. The fungus has no buffer.