Switchgrass Yield Response Functions and Profit-Maximizing Nitrogen Rates on Four Landscapes in Tennessee

TLDR: In this article, the authors used the analysis of variance approach by imposing a response function to find the yield-maximizing N rate for switchgrass, and the results have varied depending on spatial and temporal factors.

Abstract: Published in Agron. J. 104:1579–1588 (2012) Posted online 24 Aug. 2012 doi:10.2134/agronj2012.0179 Copyright © 2012 by the American Society of Agronomy, 5585 Guilford Road, Madison, WI 53711. All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. S is a warm-season, perennial grass species largely recognized for its potential as a herbaceous lignocellulosic biomass crop in North America (Vogel, 1996). Switchgrass can achieve high yields with minimal input management on land that is considered marginal for crop production (McLaughlin and Kszos, 2005). For instance, nutrients such as N move from the aboveground biomass into the root system following crop senescence, so harvesting after senescence minimizes the need for nutrient replacement (Parrish and Fike, 2005); however, annual applications of N fertilizer are needed to produce yields large enough to make growing switchgrass for lignocellulosic biomass economically viable (Heaton et al., 2004). Little attention, however, has been given to determining the N rate that maximizes producers’ profits. Several agronomic studies have estimated yield-maximizing N rates for switchgrass by performing an analysis of variance on the mean yield response to various N rates, and the results have varied depending on spatial and temporal factors. For example, Lemus et al. (2008) tested the switchgrass yield response to N for a 5-yr, large-field experiment in Iowa and found that yield was maximized when 112 kg N ha–1 was applied. With 3 yr of data from Oklahoma, Thomason et al. (2005) found that 448 kg N ha–1 achieved the highest annual switchgrass yield when the crop was harvested multiple times per year. Mulkey et al. (2006) conducted a 5-yr experiment on the switchgrass response to N on land enrolled in the Conservation Reserve Program (CRP) in South Dakota and concluded that the yieldmaximizing N rate was 56 kg N ha–1. Mooney et al. (2009) found that the yield-maximizing N rate for switchgrass grown for 3 yr on poorly drained soils in West Tennessee was 200 kg N ha–1 and the yield-maximizing N rate for switchgrass grown on moderately to well-drained soils was 67 kg N ha–1. Because previous studies had discounted the influence of soil quality and landscape on switchgrass production (Fike et al., 2006), the variation in yield-maximizing N rates on different soil types and landscapes found by Mooney et al. (2009) was an important contribution to the literature. Wullschleger et al. (2010) determined that more research is needed on this issue. The analysis of variance approach is useful to provide some insight into optimal N rates but excludes the possibility of the optimal N rate being between the discrete N rates in an experiment. A few studies have gone beyond using the analysis of variance approach by imposing a response function to find the yield-maximizing N rate for switchgrass. Muir et al. (2001) estimated the yield-maximizing N rate at 168 kg ha–1 for ‘Alamo’ switchgrass produced in Texas using linear and quadratic yield response functions. Vogel et al. (2002) estimated a quadratic response function for 2 yr of switchgrass data from Iowa and Nebraska and found a yield-maximizing N rate of 120 kg N ha–1. Recently, Haque et al. (2009) found the yieldand profitmaximizing N rate to be 65 kg N ha–1 using 3 yr of switchgrass yield data from Oklahoma. The three yield response functions considered by Haque et al. (2009) were the linear, quadratic, and linear response plateau. The linear response plateau function was ABSTRACT Little is known about how yieldand profit-maximizing N rates of switchgrass (Panicum virgatum L.) respond to environmental influences. The objective of this research was to determine the most suitable yield response functions and profit-maximizing N rates of switchgrass grown on four landscapes in Tennessee. Research was conducted in West Tennessee during a 7-yr period on four landscapes including: (i) a well-drained level upland (WDLU), (iii) a wellto moderately well-drained floodplain (WDFP), (iii) a moderate to somewhat poorly drained eroded sloping upland (MDSU), and (iv) a poorly drained floodplain (PDFP). The yield response functions considered were the quadratic, the quadratic-plus-plateau, the linear response plateau, and the linear response stochastic plateau. The most suitable response function for each landscape was used to calculate the profit-maximizing N rate. The linear response stochastic plateau function was the most suitable for the WDFP, WDLU, and MDSU landscapes, and the quadratic function was the most suitable for the PDFP landscape. The most suitable yield response function and the yieldand profit-maximizing N rates for switchgrass were sensitive across landscapes. Overapplication of N and a decrease in farmers’ net returns were predicted when a less suitable response function was selected to determine optimal N rates.