Abstract: During bird survey with line transect method, the estimation of bird density is strongly affected by the effective half-strip width (ESW). When the selected transect line width is larger than the ESW, the density of corresponding species would be underestimated. However, few studies have considered the variations in ESW. We estimated the ESW of 175 forest bird species from 42 families in Guangdong Province using the Distance software. We also examined species traits associated with the ESW. There were 52 species with ESW≤25 m, 60 species with 25 m<ESW≤50 m, and 63 species with ESW>50 m. The number of records of birds (obs) was the main factor affecting the choice of function combinations in the Distance software. The result of ESW was relatively stable when obs was greater than 20, and more accurate when the records was greater than 35. The results of nine kinds of function combination were more accurate and similar when records were high. Overall, the combination of hazardrate and cosine model performed better than other models. The ESW was significantly positively correlated with body length (BL) of birds (r=0.628, P<0.01). Our findings provide a baseline ESW=2.14×BL^0.9398 (unit: ESW, m; BL, cm). Furthermore, the bird’s call affected ESW. When the body length was similar, the ESW of birds often tweeting was 1.61 times larger than the ESW of birds calling infrequently. Our findings suggest that the heterogeneous ESW among birds should be taken into account in distance sampling analysis.

Key words: soil environment, sorghum, rotation cropping, microbial community, continuous cropping obstacle.