Background The distribution pattern of collapsed gully (Beng Gang) on different slope aspect can't be tested and quantitively analysed because of different slope aspect standards and acquisition methods in the previous studies. Here we aim to understand the suitability of different methods of extracting and dividing slope aspects in studying the slope direction of collapsed gullies, to verify the conclusion of the selectivity of slope direction in collapsed gully, to clarify the generality and applicability of the selectivity of collapsed slope aspects, and to quantitatively calculate the distribution of collapsed gullies in different slope aspects.

Methods The contours and the boundaries of collapsed gullies as well as the administration boundaries were digitalized in the 1:10 000 topographic map of Youtian town of Wuhua county, Guangdong province. The number and area of collapsed gullies on different slope aspects were calculated by ArcGIS and manual method. The number and area of collapsed gullies in the previous studies were re-counted. 2-slope-aspect refers to the slope aspect is classified into sunny slope and shady slope. 3-slope-aspect refers to the slope aspect is classified into sunny slope, semi-sunny slope and shady slope. 8-slope-aspect refers to the slope aspect is classified into north, northeast, east, southeast, south, southwest, west and northwest.

Results 1) Although the number and area of collapsed gullies on every slope aspect were different using two methods in Youtian town, but the distribution law of collapsed gullies was the same, the number and area of collapsed gullies on the sunny and southern slope aspect were separately more than those of the shady and north. Collapsed gullies on the sunny slopes accounted for 66.5% of the total while those on shady slopes accounted for 33.5% in 2-slope-aspect. The gullies on the sunny slope accounted for 45.0% and 16.9% on shady slope in 3-slope-aspect; the gullies on the southern slope accounted for 13.8% and 1.3% north slope in 8-slope-aspect. The area of shady slope land in the town was more than that of the sunny one, and the area of the northern slopes was larger than that of the south slopes. 2) After unifying slope aspect standard, the number and area of collapsed gullies on sunny slope were significantly more than those of the shady ones (P < 0.05), the gullies on the southern and southwest slopes were more than those on the northern ones. Collapsed gullies on the sunny slope accounted for 70.3% and 29.7% on shady ones in 2-slope-aspect; the collapsed gullies on the sunny slope accounted for 53.5% and 19.4% on shady in 3-slope-aspect; and the collapsed gullies accounted for 20.7% on southern slope, 20.2% on southwest slope and 5.2% on northern in 8-slope-aspect.

Conclusions The slope aspect selection law of collapsed gully in Youtian town of Wuhua county was no relevant to the slope aspect standard or computing method or the slope aspects distribution of the town land. Combing all former studies data after unifying slope standard, collapsed gullies were mainly on sunny slope, the collapsed gullies on south and southwest slopes were the most, and the smallest on north. The quantitative distribution of collapsed gullies on different slope aspects was achieved.