Abstract
Secondary succession is occurring worldwide as ecosystems regenerate following human and natural disturbances such as logging, agriculture, fires, and hurricanes. Most of today’s forests are in a state of secondary succession (as opposed to old growth), and agricultural abandonment is a significant driver of this trend. Secondary succession is currently taking place throughout the East Coast of the United States due to agricultural clearing and subsequent abandonment as agriculture has moved westward in the 19th and 20th centuries. The consequences of this shift in terms of carbon (C) cycling are relatively well understood, as the C cycle consists predominantly of a simple balance between photosynthesis and ecosystem respiration. However, the effects of large swaths of secondary successional land on nitrogen (N) cycling are less well understood, and most current empirical studies focus simply on N stocks rather than on the numerous N transformations and fluxes that occur within ecosystems. In this dissertation, I assess the current state of knowledge regarding N cycling following agricultural abandonment worldwide in a systematic literature review (Chapter 2), and I examine various changes in N stocks, transformations, and fluxes in a secondary successional chronosequence in Virginia (Chapters 3 and 4). I also explore the implications of the various stages of successional land on estimating foliar N traits using remote sensing (Chapter 4).
In Chapter 2, I find that soil N is generally depleted following agricultural land management, and increases or stays constant as succession proceeds. I also determine that soil N transformations often accelerate through succession, and that NH4+ and NO3- stocks are often inversely related. While there are numerous studies examining N stocks, and some examining N transformations, very few measure N fluxes through succession, so trends are difficult to discern. Due to the absence of studies measuring multiple aspects of N cycling at once, I conducted my own study at Blandy Experimental Farm (Chapter 3) looking at changes in several N cycling variables through successional time using two replicate chronosequences. I find evidence of increased N availability and accelerated N cycling later in succession at Blandy, with upward-trending soil N, increased N mineralization and nitrification rates, increased foliar N concentration (among and within species), and high 15N late in succession, indicating a relatively open N cycle. As foliar N is a key integrator of ecosystem N cycling, I focus on foliar N in Chapter 4, and find that foliar N concentration and leaf mass per area (LMA), an important indicator of plant strategy, change significantly through succession. Due to growing interest in the estimation of ecosystem N cycling parameters with the use of remote sensing as opposed to typical on-the-ground field methods, I also explore the potential of using leaf reflectance on the electromagnetic spectrum to estimate foliar N and LMA across successional stages. I find that models estimating LMA perform well both within and across succession, but that models predicting foliar N perform less well when making estimations across successional stages.
While N cycling through secondary succession is less well studied than C cycling, I concludex that there are many distinct changes in N cycling as succession proceeds, with potentially vast implications. I also find that common remote sensing techniques may not accurately depict foliar N traits across successional stages. Secondary successional land comprises a significant amount of the land mass worldwide, and changes in N cycling can have cascading effects across systems including changes in primary productivity, health of water bodies, and air quality. Thus, I argue that the consideration of land use history and successional dynamics should be an integral part of forest ecosystem studies, and that these factors merit further research.
