Abstract
Oak stands are declining in many regions of southern Europe. The goal of this paper is to assess this process and develop an effective monitoring tool for research and management. Long-term trends of the Normalized Difference Vegetation Index (NDVI) were derived and mapped at 30-m spatial resolution for all areas with a stable land cover of cork oak (Quercus suber L.) and holm oak (Quercus ilex L.) forests and agroforestry systems in mainland Portugal. NDVI, a good proxy for forest health and productivity monitoring, was obtained for the 1984–2017 period using Landsat-5 TM and Landsat-7 ETM+ imagery. TM values were adjusted to those of ETM+, after a comparison of site-specific and literature linear equations. The spatiotemporal trend analysis was performed using only July and August NDVI values, in order to minimize the spectral contribution of understory vegetation and its phenological variability, and thus, focus on the tree layer. Signs and significance of trends were obtained for six representative oak stands and the whole country with the Mann Kendall and Contextual Mann-Kendall test, respectively, and their slope was assessed with the Theil-Sen estimator. Long-term forest inventories of six study sites and NDVI time series derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) allowed validating the methodology and results with independent data. NDVI has a good relationship with cork production at the forest stand level. Pettitt tests reveal significant change-points within the trends in the period 1996–2005, when changes in drought patterns occurred. Twelve percent of the area of oak stands in Portugal presents significant decreasing trends, most of them located in mountainous regions with shallow soils. Cork oak agroforestry is the most declining oak forest type, compared to cork oak and holm oak forests. The Google Earth Engine platform proved to be a powerful tool to deal with long-term time series and for the monitoring of forests health and productivity.
Highlights
Cork oak (Quercus suber L.) and holm oak (Quercus ilex L.) cover one-third of the woodland areas in continental Portugal
Since the populations do not follow a normal distribution and time series observations are auto-correlated by nature, the percentages of mean and median differences between raw and adjusted Normalized Difference Vegetation Index (NDVI) values were compared to determine the best relation to use for Portugal oak areas
Based on previous works exposed in the Introduction section regarding intra-annual variations of NDVI in woodlands, only July and August NDVI values were used for the trend analysis
Summary
Cork oak (Quercus suber L.) and holm oak (Quercus ilex L.) cover one-third of the woodland areas in continental Portugal. Several studies revealed disparities between the sensors, in particular between Landsat TM 5 and ETM+ 7 NDVI values that can introduce artificial long-term trends, due to differences in the bands wavelength sensibility and atmospheric conditions [28,30,31]. Our study will rely on inter-calibrated Landsat-derived NDVI time series, and will use MODIS data to validate the adjustments performed Phenological monitoring of both Californian Quercus douglassii savannas and pure grasslands areas with NDVI [32] showed phenological dates were consistent across spatial resolution for grasslands but varied in savannas. We developed a flexible methodology of long-term monitoring based on remotely sensed data that can be applied to other types of forests and vegetation indices
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
