2026 Volume 52 Issue 1 Pages 49-54
A complete inventory of 325 street trees was conducted along four streets in Chuo ward, Kobe, and six ecosystem-service metrics—tree replacement value, carbon storage and sequestration, air-pollutant removal, stormwater-runoff reduction, and annual total benefits—were estimated using i-Tree Eco. Bootstrap resampling simulations were then applied to five datasets defined by street section and dominant species to examine required sample size and estimation uncertainty. Estimated values differed among streets and species, reflecting variation in species composition, tree size, and crown condition. The required sample size satisfying the 10% relative mean error criterion ranged from 6 to 39 trees. For annual total benefits, the relative confidence-interval width remained approximately 48% even at the required sample size. These results suggest that, in practical applications, survey sample sizes should be determined according to the purpose of evaluation, and estimates should be reported together with their uncertainty.