To study the diversity of ground-dwelling beetle communities and their environmental relationships, and to understand their efficacy as habitat indicators and sensitivity to habitat perturbations, we used pitfall traps to investigate patterns of ground-dwelling beetle diversity along ecotones between the desert hinterland and desert-oasis ecotone along a desert edge in the southern Gurbantunggut Desert, including a disturbed habitat by desert highway protecting system. Our main results were as follows: (1) In total, we trapped 54,527 individuals of 81 species, belonging to 14 families, among which Tenebrionidae and Curculionidae were most dominant and accounted for 93.65% and 5.14% of abundance, respectively. Tenebrionidae was most abundant in the desert hinterland, while Curculionidae dominated ecotones both in terms of richness and abundance. All Platyope spp. showed a preference for aeolian sandy soil, while Adesmia aweiensis and Trigonoscelis sublaevigata sublaevigata showed a preference for alkaline desert soil. (2) Species richness, activity density, diversity, dominance, and evenness indices all showed difference among habitats (df = 32, P<0.05). (3) DCA (detrended correspondence analysis) ordinations of habitats based on ground-dwelling beetle communities or dominant groups showed that desert edge was the transitional area in our study area, and that communities on slopes were more similar to those in desert hinterland while those on interdunes were more similar to those in ecotones. (4) Factors including species richness and diversity, evenness and dominance indices were linearly correlated (P<0.05) with soil water content, nutrient content, total N, and total P content. Among them, the correlation with soil water content was the highest, and activity density showed the lowest and negative correlations with all environmental factors. (5) A significant decline in beetle activity density was detected within the highway protecting system (P<0.05); species richness, diversity index and evenness index all declined in the same area, although the differences were insignificant. The relative importance of main species also changed. Conclusions we can draw based on these results are: These communities show low richness, diversity, and evenness but high dominance and abundance. This becomes more pronounced as one moves from the desert-oasis ecotone to the desert hinterland. Desert ground-dwelling beetles showed certain habitat-indicative value at the family, genus and species levels. Soil type is likely the primary mechanism leading to the observed community ordination results. The negative relationship between activity density and soil water content might be a general rule of desert ground-dwelling beetle’s distribution at local scales. The desert highway protecting system within our study results has affected ground-dwelling beetle communities, and associated artificial plants (straw barrier, Haloxylon ammodendron and Calligonum sp.) might be the driving factor leading to these changes.