Coffee Science, Lavras, v. 14, n. 2, p. 131 - 137, apr./jun. 2019 Coltri, P. P. et al. 131 COFFEE LAND COVER CHANGES ANALYSES: A STUDY CASE IN SÃO PAULO STATE Priscila Pereira Coltri 1 , Paulo Renato Lima 2 , Andrea Koga-Vicente 3 , Renata Ribeiro do Valle Gonçalves 4 (Received: January 10, 2019; accepted: April 15, 2019) ABSTRACT: In this research, we combine data analyses with hotspots method to identify the spatio-temporal trend of São Paulo’s coffee cultivation area. Our hypothesis is that coffee cultivation area has been changing signifcantly in the study area since 1990. Therefore, the main goal of this research was to map the spatial pattern of coffee land use change. For coffee land use diagnostics, offcial data of cultivated area, hotspot analyses and growth rate were used. The results demonstrated that coffee cultivation area decreased and concentrated in smaller areas, which are traditionally recognized as “coffee quality regions”. The producer size analyses evidenced that, not only the localization, but also the producer profle changes as well. Smallholders increased but medium and large producers decreased signifcantly in the studied period. The coffee abandonment analyses demonstrated that, over the study period, 51.46% of the coffee area cultivated in the study region was abandoned. Index terms: Land cover changes, hotspots, coffee abandonment. MUDANÇA DO USO DO SOLO CAFEEIRO: O ESTUDO DE CASO DO ESTADO DE SÃO PAULO RESUMO: Nesta pesquisa, combinou-se análise de dados com o método de hotspots para identifcar a tendência espaço- temporal da área de cultivo de café no estado de São Paulo. A hipótese é que a área cultivada com café vem se alterando signifcativamente no Estado desde 1990. Assim, o objetivo principal da pesquisa foi mapear o padrão espacial das mudanças no uso do solo cafeeiro. Para o diagnóstico do uso do solo do café, foram utilizados dados ofciais da área cultivada, análises de hotspots e taxa de crescimento. Os resultados demonstraram que a área cultivada de café diminuiu e se concentrou em pequenas áreas, que são regiões tradicionalmente reconhecidas por sua qualidade. As análises do tamanho do produtor evidenciaram que, não apenas a localização, mas também o perfl do produtor mudou. O número de pequenos agricultores aumentou no período estudado, enquanto os produtores médios e grandes diminuíram signifcativamente. As análises de abandono de área demonstraram que, ao longo do período estudado, 51,46% da área destinada ao plantio de café foi abandonada na região de estudo. Termos para indexação: Mudança do uso do solo, hotspots, abandono de áreas cafeeiras. 1 INTRODUCTION According to Smith (1985), coffee cultivation has begun in AD575 and the frst documentation about this crop as we know it is by Razes, from the 10 th century. Since then and over the centuries, coffee cultivation has become popular, turning into one of the most consumed beverages in the world. Brazil has dominated the position of largest coffee producers for over 150 years (ICO, 2018). Historically, São Paulo is a traditional coffee producer. However, this position has been lost over time, according to offcial data from the Brazilian Institute of Geography and Statistics. Land use and land use changes (LULCC) are a result of complex interactions that involve social, ecological, political, industrial and economic drivers (KLEEMANN et al., 2017). 1 Universidade Estadual de Campinas/UNICAMP) - Centro de Pesquisas Meteorológicas e Climáticas Aplicadas a Agricultura/ CEPAGRI - Cidade Universitária “Zeferino Vaz” - 13.083-970 - Campinas/SP- pcoltri@unicamp.br, renata@unicamp.br 2 Universidade Estadual de Campinas/UNICAMP - Instituto de Geociências/ IG - Rua Carlos Gomes, 250 - Cidade Universitária 13.083-896 - Campinas - SP - paulorenato510@gmail.com 3 Embrapa Meio Ambiente - Plataforma ABC - Rodovia SP 340, Km 127,5, S/N - Tanquinho Velho - 13.820-000 - Jaguariúna - SP andrea.kvicente@gmail.com Depending on the magnitude and dimension of these interactions, the result may be deeper than cultivation migration and include crop abandonment (SAMPER, 2010). Assessing spatial changes pattern on adequate spatial data is essential to understanding future consequences and land system dynamics (KUEMMERLE et al., 2016). To analyze change pattern in the agriculture landscape, at different scales, remote sensing techniques are frequently used (GONÇALVES et al., 2012). For temporal evaluation, low spatial with high temporal resolution images are indicated because of the frequency of the data (ZULLO JUNIOR et al., 2014). However, according to Nogueira et al. (2015), recognizing coffee crops changes in low spatial and high temporal resolution images is not trivial because of the crop features and the planting