In 1904, a necrotrophic fungus called chestnut blight began devastating North America’s native chestnut trees. More than a century later, a grove of approximately 200 blight-resistant American chestnut trees in West Virginia is facing the threat of being cleared to make way for a large-scale artificial intelligence data center. This fungus, brought across the Pacific on chestnut trees from Japan and China, attacks tree bark and kills the above-ground portions of susceptible trees. The disease eventually wiped out billions of chestnut trees across the US.
Planted in the 1970s, the 56-year-old trees are a valuable source of genetic material that could help restore the species. Fewer than 4,300 acres of fruit-producing chestnut trees existed across the continental US in 2023, according to the supplied report.
The project is planned to span more than 2,250 acres and provide 8 gigawatts of computing capacity, with its developer calling it “one of the largest single AI infrastructure investments in US history. ” Researchers and horticulturists have since worked to preserve the species through breeding programmes and blight-resistant hybrids. However, the proposed Monarch Compute Campus, developed by Nscale in collaboration with Nvidia and Caterpillar, could put the grove at risk. However, mature chestnut groves remain uncommon.
Under the current plans, the completed campus will cover more than 2,250 acres and provide 8 gigawatts of computing capacity for AI companies, including Anthropic. Losing the grove could therefore affect more than the existing trees, potentially reducing the resources available to future restoration projects.
The surviving chestnut trees are “irreplaceable” because they could support future restoration work, according to From the Front Lines. The trees could provide seeds and genetic material for efforts to establish new chestnut populations. The proposed data center development is being built by Nscale in collaboration with Nvidia and Caterpillar. The development represents a substantial investment in computing infrastructure as technology companies expand their capacity to train and run artificial intelligence models.