Time series analysis of long-term vegetation trends, phenology, and ecosystem service valuation for grasslands in the U.S. Great Plains

Abstract

Grasslands are one of the largest, most biodiverse, and productive terrestrial biomes but they receive very low levels of protection. The temperate grasslands in the United States are one of the most threatened grassland ecosystems. Every year, a significant portion of grasslands in the Great Plains are converted to agricultural use, with almost 96% of the historical extent lost. Other factors that affect existing grassland health include significant climatic changes, invasion of woody, non-native species, fragmentation, lack or inadequate burning, and excessive grazing. The impact of the combination of these factors on grasslands in the US Great Plains is still unknown. The goal of this research is to investigate the long-term grassland vegetation conditions using a well-known indicator (greenness) and assesses its impact on the provision of select grassland ecosystem services within the US Great Plains from 2001 to 2017. The above goal was achieved with three objectives addressed in three chapters. In Chapter 3, a time-series analysis of Moderate Resolution Imaging Spectrometer (MODIS) 16-day maximum value composite Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) data (MOD13Q1 Collection 6) was performed to assess long-term trends in vegetation greenness across the Great Plains ecoregion of the United States. The Breaks for Additive Season and Trend (BFAST) decomposition method was applied to a time series of images from 2001 to 2017 to derive spatially explicit estimates of gradual interannual change. Results show more 'greening' trends than 'browning' and 'no change' trends during the study period. Comparing the trend results from both vegetation indices suggests that EVI is more suitable for this analysis in the study area, especially in areas with high biomass. In Chapter 4, a time-series analysis of Moderate Resolution Imaging Spectrometer (MODIS) 16-day maximum value composite Enhanced Vegetation Index (EVI) data (MOD13Q1 Collection 5) is used to explore spatial patterns of vegetation phenology and to assess long-term phenology trends across the region. The program TIMESAT was used to extract key measures of vegetation phenological development from 2001 to 2017, including the phenometrics (1) season length, (2) start of growing season, (3) end of growing season, (4) middle of the growing season, (5) maximum NDVI value, (6) small integral, (7) left derivative, and (8) right derivative. Results show important variation in phenological patterns across the region such as a shift to a later start, earlier end, and shorter the growing season length, especially in the southern parts of the region. As shown in the small integral and maximum EVI, vegetation productivity appears to have increased over many areas in the Great Plains ecoregion. Finally, Chapter 5 focuses on developing a methodological improvement to the widely used Invest ecosystem services model that uses remotely sensed inputs to capture the interannual spatio-temporal dynamics of grassland vegetation on the provision of grassland ecosystem services across the US Great Plains. A selected set of grassland ecosystem services was quantified (economic and biophysical values) for the period between 2001 and 2017. This exploratory study will be a basis for highlighting the role grasslands play in providing essential ecosystem services and how improved long-term vegetation monitoring can benefit land-use decisions locally and regionally.