Impact of cultivated land fragmentation on spatial heterogeneity of agricultural agglomeration in China

Systematically revealing the impact of cultivated land fragmentation(CLF)on the geographical agglomeration pattern of agricultural specialization(AS)has positive significance for national agricultural production management.Based on the data of the second national land survey and agricultural production,this study has explored the impact of CLF on spatial heterogeneity of agricultural agglomeration in China by comprehensively using the Theil index,ordinary least square model and geographically weighted regression.Results showed that:(1)the regional differentiation of the CLF in China is obvious,and the cultivated land fragmentation index is generally characterized by increasing pattern from northwest to southeast.(2)Spatially,the development level of AS in China has formed three high-value clusters in the Northeast China Plain,the Qinghai-Tibet Plateau,and the middle of the Middle-lower Yangtze Plain;and the low-value contiguous areas centered on the Yunnan-Guizhou Plateau and the Sichuan Basin and surrounding regions,with significant spatial differences.The contribution of grain crops,economic crops,and vegetables and melon to the level of AS was 74.63%,9.09%,and 16.28%,respectively,and the pattern of agricultural geographical aggregation dominated by grain crops has primarily taken in shape.(3)CLF is significantly negatively correlated with AS,and every 1%increase in the degree of CLF will result in a decrease of about 0.2%in AS.However,the impact of CLF on the geographic agglomeration of different crop categories or groups varies significantly.Among them,CLF has a prominent impact on the specialization level of grain crops and vegetables and melon.Each 1%increase in the CLF will reduce the specialization level of grain crops by 0.38%,and increase the level of vegetables and melon by about 0.22%.(4)According to the landscape characteristics of cultivated land,the degree of spatial division and agglomeration of cultivated land patches have a significant impact on the formation of geographical agglomerationpattern of AS,and the intensity and direction of influence show significant regional differentiation,while the patch size has no significant impact.

Design of a spatial sampling scheme considering the spatial autocorrelation of crop acreage included in the sampling units

Information on crop acreage is important for formulating national food polices and economic planning. Spatial sampling, a combination of traditional sampling methods and remote sensing and geographic information system （GIS） technology, provides an efficient way to estimate crop acreage at the regional scale. Traditional sampling methods require that the sampling units should be independent of each other, but in practice there is often spatial autocorrelation among crop acreage contained in the sampling units. In this study, using Dehui County in Jilin Province, China, as the study area, we used a thematic crop map derived from Systeme Probatoire d＇Observation de la Terre （SPOT-5） imagery, cultivated land plots and digital elevation model data to explore the spatial autocorrelation characteristics among maize and rice acreage included in sampling units of different sizes, and analyzed the effects of different stratification criteria on the level of spatial autocorrelation of the two crop acreages within the sampling units. Moran＇s/, a global spatial autocorrelation index, was used to evaluate the spatial autocorrelation among the two crop acreages in this study. The results showed that although the spatial autocorrelation level among maize and rice acreages within the sampling units generally decreased with increasing sampling unit size, there was still a significant spatial autocorrelation among the two crop acreages included in the sampling units （Moran＇s / varied from 0.49 to 0.89）, irrespective of the sampling unit size. When the sampling unit size was less than 3000 m, the stratification design that used crop planting intensity （CPI） as the stratification criterion, with a stratum number of 5 and a stratum interval of 20% decreased the spatial autocorrelation level to almost zero for the maize and rice area included in sampling units within each stratum. Therefore, the traditional sampling methods can be used to estimate the two crop acreages. Compared with CPI, there was still a strong spatial correlation among the two crop acreages included in the sampling units belonging to each stratum when cultivated land fragmentation and ground slope were used as stratification criterion. As far as the selection of stratification criteria and sampling unit size is concerned, this study provides a basis for formulating a reasonable spatial sampling scheme to estimate crop acreage.