樟白輪蚧（Aulacaspis yabunikkei）在樟樹上之空間分布及取樣技術

Spatial Distribution and Sampling of Aulacaspis yabunikkei (Homoptera: Diaspididae) in Camphor Trees

比較樟白輪蚧（Aulacaspis yabunikkei Kuwana）各齡期在樟樹上株間、枝間及葉間的Iwao's patchiness regression 及Taylor's power law 的空間分布指數及決定在樟樹上的取樣單位及最適取樣齡期。結果以Taylor's power law 的相關性較高（R^2：0.89 - 0.99），Iwao's patchiness regression 則較差（R^2：0.54 - 0.90）。各齡期在株間、枝間及葉間的Taylor's power law 及Iwao's patchiness regression 聚集指數範圍分別為1.76 - 2.65 及1.26 - 11.83，顯示各齡期在樟樹上的分布都屬聚集型（b ＞ 1.7 及 β ＞ 1.2）。樟白輪蚧在樟樹上內層的分布顯著較外層多，在老葉上分布較嫩葉上多，且葉背的數目也較葉面上多。利用樟樹上分層取樣的結果，各齡期和總蟲數的相關性以卵為最高（R^2 = 0.89，P ＜ 0.01），適合推薦專業人士取樣估計總蟲數。不過因卵太小難以肉眼觀察計數，推薦以株內枝條上老葉上的雌介殼數（和總蟲數的相關性，R^2 = 0.68，P ＜ 0.01）為第一線防治人員取樣的對象。在樟白輪蚧嚴重危害時（每株100 隻時），如只要求0.5 的精密水平時，則只需取樣3 株即具代表性。

"Dispersion patterns of varied stages generated by Iwao's patchiness regression and Taylor's power law for Aulacaspis yabunikkei Kuwana in different sampling units of camphor trees, Cinnamomum camphora（L.）, were compared. Taylor's power law provided a consistently good fit to the data, whereas the fit of Iwao's patchiness regression were erratic, and the values of aggregation index of Taylor's power law（1.76 to 2.65）were narrower than those of Iwao's（1.26 to 11.83）, but both indices（b＞1.7 andβ＞1.2）indicate a clumped distribution pattern in all sampling units. Mean numbers of scales per leaf differed significantly（P＜0.05）between the lower and upper layer of the canopy, between old and young leaves, and between the underside and upper surface of leaves. The number of eggs corresponded the most closely to the total population, and this stage is best for precisely estimating the population. However, for the non-professional, the eggs are difficult to count. The number of female adult scale covers were easy to count and also corresponded very closely to the total population. Thus, female adult scale covers on the under leaf surfaces of old leaves were chosen as sampling targets to represent the scale population and for the decision making of pest management. Scales only with high density would damage camphor tree heavily, therefore, in fixed sampling plan, when there were 100 female scales on underside of leaf per twig in a tree, the optimum samples of 12 twigs per tree and 3 twigs per tree would reach 0.25 and 0.5 precision levels, respectively."