Analysis of Litterbag Cover Across Drylands

Arid and semiarid ecosystems (known as drylands) cover 45% of global land area and are predicted to expand to encompass half of the world’s land area by the end of the century. Litter decomposition plays a large role in nutrient and carbon cycling in dryland ecosystems, yet it remains poorly understood. Models that accurately predict decomposition in mesic ecosystems fail to accurately describe decomposition in drylands due to differing drivers of decomposition but also because litter in drylands accumulates around litter retention elements (LREs). LREs can be any object or surface that inhibits the movement vectors (e.g., wind) that push litter across drylands, creating a “pool” of litter around the LRE. Litter pooling increases the amount of mixing between litter and soil, creating a microclimate more conducive to microbial decomposition. Due to the increase in microbial decomposition, the decomposition rate for litter around LREs can be markedly different than that of litter not in LREs. To further understand how much litter accumulates in LREs, I studied the differences in litter accumulation between LREs and open areas in five drylands across the Southwestern United States. To do this, I visually analyzed photos of 424 litterbags to determine the cover percentages of four different types of organic litter (grass, broadleaf, reproductive, woody) and rock litter. Visual analysis of litterbags consisted of manually delineating the percent coverage of each of these litter categories. Litterbags had been placed in both open intercanopy areas as well as woody sub-canopy areas in which the plant canopy acted as the LRE. Additionally, 45 of these litterbags were randomly selected for analysis in the computer program FIJI (FIJI is Just ImageJ) to assess the litter area find the percent difference between visual and digital analysis. Areas underneath woody sub-canopies accumulated far more organic matter litter over time than open areas between canopies did but displayed a similar amount of rock litterbag cover. Shrub microsites also displayed far more varied litterbag cover percentages than open microsites. Data also suggested that litter does not always accumulate underneath shrubs or open intercanopy areas and may dissipate as time progresses. These results support the idea that litter accumulation varies throughout drylands, and that soil and litter mix frequently in LREs such as under woody plant canopies. The percent difference between FIJI analysis and visual analysis was generally negative, reflecting that visual estimation of litterbag cover was typically smaller than digital estimates. Cumulatively, litter was shown to accumulate much more around LREs and even move from them – supporting the idea that litter decomposition models need to account for litter movement in drylands to be accurate.