Influences of nitrogen fertilization and climate regime on the above-ground biomass yields of miscanthus and switchgrass: A meta-analysis

Highlights

• Switchgrass was more responsive to N fertilization than miscanthus.

• Miscanthus biomass increased to N addition rates, but response decreased with years.

• Excessive N addition rates and duration did not further increase switchgrass yields.

• N effects on switchgrass depend on annual mean temperature and precipitation.

Abstract

Perennial grasses are touted as sustainable feedstocks for energy production. Such benefits, however, may be offset if excessive nitrogen (N) fertilization leads to economic and environmental issues. Furthermore, as yields respond to changes in climate, nutrient requirements will change, and thus guidance on minimal N inputs is necessary to ensure sustainable bioenergy production. Here, a pairwise meta-analysis was conducted to investigate the effects of N fertilization (amount and duration) and climate on the above-ground biomass yields of miscanthus (Miscanthus x giganteus) and switchgrass (Panicum virgatum L.). Both regression models and meta-analyses showed that switchgrass was more responsive to N than miscanthus, although both showed significant and positive N effects. Meta-analysis further showed that the positive growth response of miscanthus to N application increased with N addition rates of 60–300 kg N ha⁻¹ year⁻¹, but the magnitude of the response decreased with the number of years of fertilization (duration). N effects on switchgrass biomass increased and peaked at rates of 120–160 kg N ha⁻¹ year⁻¹ and 5–6 years of N inputs, but diminished for rates >300 kg N ha⁻¹ year⁻¹ and >7 years. Meta-analysis further revealed that the influences of N on switchgrass increased with both mean annual temperature and precipitation. Miscanthus yields were less responsive to climate than switchgrass yields. This meta-analysis helps fill a gap in estimation of biofeedstock yields based on N fertilization and could help better estimate minimum N requirements and soil management strategies for miscanthus and switchgrass cultivation across climatic conditions, thereby improving the efficiency and sustainability of bioenergy cropping systems.

Introduction

Bioenergy crops have been increasingly cultivated across the globe since early 1990s [1] as a result of policies seeking the increased use of biofuels as part of a renewable energy portfolio, such as the goals of 10% of transportation fuel has been set for the European Union by 2030 [2] and 12% in the United States by 2025 [3]. In part due to goals and mandates such as these, global biofuel demand is projected to increase to about 760 million tonnes of oil equivalent by 2050, which would fulfill 27% of total demand for transportation fuels [4]. To meet this future demand, perennial grass species such as miscanthus (Miscanthus x giganteus) and switchgrass (Panicum virgatum L.) are promising agronomic options. Perennial grasses have advantages for water quality, soil quality, and other dimensions of environmental sustainability, and due to their broad tolerance for initial soil conditions, they can be grown on underutilized land with poor or degraded soils. This is important because it allows currently productive cultivated lands to remain in food production [5]. It has also been suggested that sustainable cultivation of perennial bioenergy crops, such as these, could increase soil carbon (C) sequestration and thereby further offset overall C emissions and consequences thereof [6].

The environmental benefits of bioenergy crops, such as switchgrass and miscanthus, may be offset if excessive fertilizers are applied for bioenergy crop production [7]. Consequences can include nitrogen (N) leaching to watersheds and increased soil emissions of nitrous oxide (N₂O), a potent greenhouse gas and ozone depleting substance. Jørgensen et al. [8] reported that soils cultivated with perennial bioenergy crops produced twice the amount of N₂O than conventional crops in arable soil, especially after fertilization. Thus, only when care is taken to apply the minimal amount of fertilizer required for crop growth, can biofeedstock production achieve carbon neutral or positive carbon sequestration benefits.

Globally, the yield response to N inputs by the two bioenergy crops considered here have been well documented, but results vary widely with soil management strategies and across climatic conditions. Cadoux et al. [7] reviewed a total of 27 miscanthus studies in Europe and the US dealing with above-ground dry matter production in response to nutrients. Both negligible and significantly increased yield responses to N inputs were reported due to high variation in the initial soil mineral N content. The authors also suggested that the high N losses from fertilized miscanthus were due to excessive and improperly timed applications of N fertilizer. Heaton et al. [9] compiled data from 21 peer-reviewed articles, representing 174 observations to compare responses of biomass yield to water, temperature, and N for miscanthus and switchgrass. Their study results showed that both grasses had a significant positive response to growing season precipitation and N, but not to growing degree days. In addition, Miguez et al. [10] examined the effects of soil management on the biomass production of miscanthus in Europe to assess the relationship between harvest time and dry biomass production. Using non-linear mixed models, their analyses revealed that biomass yields only responded to N fertilizer after the third growing season. Meanwhile, Wullschleger et al. [11] reviewed 39 switchgrass field trials in the USA. A parametric yield model was constructed to relate biomass yields to N fertilization, growing season precipitation, and annual temperature for two switchgrass ecotypes, and their models explained one-third of the total observed data variation. A spatially explicit modeling effort has also shown different responses of the two switchgrass ecotypes, with strong responses to precipitation [12]. As many field trials for switchgrass occurred at relatively few agricultural field stations, this study assumed spatial correlation within, but not among, sites [12]. The studies above were published between 2007 and 2010, and no synthesis has been performed since.

One of the purported strengths of these earlier syntheses was the ability to draw generalized conclusions across geographic regions, however their power to detect effects was also limited by combining disparate studies across sites with widely varying null conditions. Thus, a central challenge remains to account for the experimental variability between studies, including among locations and years with very different climatic norms. For example, to experimentally address this problem, Laurent et al. [13] took advantage of a reference crop grown at the same sites with the tested bioenergy crops, and made indirect comparisons of the yields of different species grown at different sites. However to our knowledge, there has been no comprehensive review and synthesis regarding bioenergy crop yields responding to N levels and climate variation based on meta-analysis techniques and using global studies. This information is urgently needed to make economic and ecological management decisions that reduce economic and environmental costs by minimizing nutrient loss in bioenergy cropping systems. Pairwise meta-analysis methods can account for the heterogeneity originating from different experimental designs and null conditions as this approach directly compares the effect sizes produced by pairwise comparisons of treatment and control within studies [14] rather than overall means.

In this study, a pairwise meta-analysis approach was used to assess the effects of N fertilization and climate regime on the above-ground biomass yields of miscanthus and switchgrass in comparison with linear and non-linear regression models that have been applied in the past. Specifically, this study addressed the following questions: (1) Do miscanthus and switchgrass yields respond positively to increasing N levels and number of years of application (duration)? (2) How does mean annual temperature and total annual precipitation influence N effects on bioenergy crop yields? (3) Compared to regression-based approaches, does a pairwise meta-analysis provide a more sensitive approach to detecting responses of bioenergy crop yields to N across climatic regimes?