| Over the past decades, there has been growing evidence that urban sprawl has spread rapidly around China and even the world with fast economic development and population growth. In response, some ecological environmental problems associated with urbanization has influenced the natural phenomena and ecological processes of earth surface system on different scales from urban, regional to global, and has caused land use and land cover(LULC) changed greatly, which will directly affect the net primary productivity(NPP) of vegetation. Study on urban expansion and LULC changes and their relative influences on NPP will help us better understand the feedback effects of urban ecological structures and functions on natural environment, which has an important significance for protecting urban ecological environment. Guangzhou is chose as the study area in this study. The urban expansion characteristics and LULC changes are monitored and analyzed, spatiotemporal distributions and dynamic changes of interannual NPP and seasonal NPP from 2001 to 2013 are analyzed based on the Carnegie-Ames-Stanford Approach(CASA) model and the influences between climate change, urbanization and subsequent LULC change on NPP are also explored. The main study results and innovations are as following:(1) Urban land increases rapidly and land cover changes significantly in Guangzhou during the study period. Urban lands increase by 1512.24 km2 with an annual growth rate of 11.3%, and experience four stages of urban growth: low rates from 1979 to 1990, increased rates from 1990 to 2001, high rates from 2001 to 2009 and steady increased rates from 2009 to 2013. Other land types are influenced by urban expansion and croplands reduces significantly by 1746.04 km2, and about 769.46 km2 is converted into urban land during the 35-year study period. Other fertile lands, like forests and cropland/natural vegetation mosaics are severely impacted as a result, and the landscape becomes heavily fragmented.(2) Declining trends occur in Guangzhou and its five regions during the period from 2001 to 2013 on a time scale. Mean NPP decreases significantly(p < 0.05)with an annual variation trend of-8.57 NPP g C/m2 and a total loss of 0.38 Tg C in study area. The highest and lowest value of mean NPP occurs in the northeastern region(1067.31 g C/m2) and the central region(292.06 g C/m2) respectively, and the mean NPP decreases extremely significant(p < 0.01) in the central, southern and northern regions.A distinct difference of mean NPP is found in spatial distribution, and the interannual NPP changes are significant. The mean NPP change of the three transects shows an obvious geographic heterogeneity from the centre to north, the centre to south and the centre to east. The highest value of mean NPP appears in the centre to north transect(431.76 g C/m2), the lowest value appears in the centre to south transect(337.53 g C/m2), and the mean NPP of blocks of every transect presents an overall downward trend from east, south and north to centre. Most pixels show a reduced trend in different study periods expect the period of 2005-2009. The variation coefficients(C.V.) of mean NPP are more significant at the pixel scale in the central, northern and southern regions during the 13 years. The spatial pattern of variation slope and significant level of slope are analyzed, the results show that NPP reduces in all regions and especially in the northern region with a mean slope value of-14.98%, followed by the southern region with-8.44%. About 28.3% of pixels of the extremely significant decrease(p < 0.01) and significant decrease(p < 0.05) are mainly in the northern region, and the pixels of extremely increase(p < 0.01) and significant increase(p < 0.05) are concentrate in the northeastern region mainly consist of forest lands.(3) The significant differences of mean NPP are found in different seasons and land cover types. The highest value of mean NPP occurs in summer of 454.39 g C/m2 and the lowest occurs in winter of 54.08 g C/m2. Mean NPP reduces significantly in summer with a variation trend of-4.18 g C/m2, while with-0.99 g C/m2 in winter. At the pixel scale, mean NPP shows a decreasing trend in all seasons and reduces faster in spring and winter than summer and autumn based on the NPP slope changes. The highest and the lowest value of mean NPP appear on evergreen broadleaf forests(1327.22 g C/m2) and shrub-grassland(483.08 g C/m2) respectively. Almost all vegetation types have the highest values of mean NPP in the northeastern region, and have the lowest values in the central region.(4) Urbanization has a significant impact on NPP changes. The changes of mean NPP are analyzed in different buffers, and the lowest value occurs in 5 km buffer(179.41 g C/m2) and the highest occurs in > 70 km buffer(1225.87 g C/m2), and mean NPP increases extremely significant(p < 0.001) in different distance buffers with the increase of distance from the city centre. Negative correlations between NPP and human activities scope and intensity are found in different buffers, human activities intensity increases by 0.1802 in 10-40 km buffer from 2001 to 2013, while the mean NPP decreases significantly in this buffer. For mean NPP changes of all land cover in different buffers, the highest value appears in > 40 km buffer and shows an obvious decreasing tendency and the lowest value appears in 5 km buffer.(5) Land cover change caused NPP changed greatly. The land covers of transfer-in and transfer-out play important impacts on NPP change, and the transformation of land covers results in a reduction of total NPP in different periods expect the period of 2001-2005. Urban expansion and croplands shrinkage are the main causes for total NPP reduction in the range of NPP decreased significantly and extremely significant, and the range of croplands in 2001 and building lands in 2013 from 2001 to 2013.The differences of NPPlulc on the spatial distribution are significant. NPPlulc has the highest value in the central region(1353.05 g C/m2) due to the increasing human interference, followed by the southern region with 1248.21 g C/m2, and the lowest value occurs in the northeastern region(521.83 g C/m2). Moreover, NPPlulc varies with different intensity of human activities, and the mean NPPlulc of blocks of every transect decreases extremely significantly(p < 0.01) from centre to east, south and north.(6) Climate change affects NPP change significantly. Positive correlations are identified between NPP and temperature and solar radiation, while negative correlations between NPP and precipitation are also found in different years, regions and land covers. For the comprehensive influences of climate variables on NPP, NPP of most pixels are mainly controlled by temperature and solar radiation, and precipitation always becomes a limiting factor for NPP accumulation. Topographic factors play important roles on forest NPP, and the impact of elevation on forest NPP is greater than aspect and slope.(7) Great differences of influence quantity and influence rate of climate change and land cover change on NPP are found. During the period from 2001 to 2013, the loss of total NPP is 0.2892 T g C and 0.4239 T g C due to climate change and land cover change respectively. Evergreen broadleaf forests are more sensitive to climate change than land cover change, while some other land cover types are greatly influenced by human activities, for example, croplands, cropland/natural vegetation mosaics and urban green lands. In the spatial scale, NPP change is mainly controlled by climate change in the northeastern region, while mainly impacted by land cover change in the northern, eastern, southern regions, and mainly impacted by interaction effects in the central of the central region. |
PDF Full Text Request