Are mixed-tree plantations including a nitrogen-fixing species more productive than monocultures?

Highlights

• 148 case studies, from 34 plantations, were inventoried from the literature.

• Mixed-tree plantations were 18% more productive than the non-N₂ fixing monocultures.

• The effect was significantly different from 0 under temperate conditions only.

• The effect was negatively correlated with biomass production in the monoculture.

• The success of the mixture seems limited to low productivity sites.

Abstract

The inclusion of N₂-fixing tree species in tree plantations has the potential to increase biomass production compared to monocultures. Both successes and failures have been described in the literature; however, it is still difficult to distinguish a general pattern and to disentangle the factors influencing the mixture effect. The first objective of this study was to provide an overview of the published data on the effect of the introduction of N₂-fixing trees in tree plantations through a meta-analysis approach and to calculate a mean effect of mixed-tree plantations on biomass production compared to monocultures of the non N₂-fixing species in stands 2–20 years of age. The second objective was to evaluate the effects of (1) climate zone (temperate vs. tropical), (2) the species used (eucalypts vs. other non N₂-fixing species, and leguminous tree species vs. other N₂-fixing species), (3) the proportion of N₂-fixing species compared to the non-fixing species, and (4) plant developmental stage. A total of 148 case studies from 34 experimental plantations under tropical (68 case studies) and temperate (80 case studies) conditions were identified from the literature. The global mixture effect was significantly positive, mixed-tree plantations being 18% more productive than the non N₂-fixing monocultures, and this effect was significantly different from zero under temperate conditions (24% more productive) but not under tropical conditions (12% more productive). Indeed, the sites where the positive mixture effect was significantly different from zero were mostly located in a temperate climate, where soil nitrogen is generally considered less available than in tropical latitudes. Intermediate and high proportions of N₂-fixing species gave similar positive results (27% more productive), while low proportions had no significant impact. Neither plantation age nor type of N₂-fixing species (legume trees vs. other N₂-fixing species) had any significant effect. In conclusion, it appears that climate is the main factor influencing the success of the mixture; however, it also seems that the degree of mixture success is more marked on sites with low biomass production where the monoculture is the least productive.

Introduction

In 2012, nearly half of all industrial round wood harvested worldwide was removed from planted forests, the majority of which were large-scale tree plantations (Payn et al., 2015). Large-scale tree plantations, most of which are located in Asia and the Americas, can occupy anywhere from hundreds of hectares to hundreds of thousands of hectares and are generally under government or commercial management (Kanowski and Murray, 2008). Such plantations often comprise a single species or a few productive, and predominantly exotic, tree species that are intensively managed for varying commercial purposes, mainly for timber and pulpwood, but also for biofuels and carbon credits (Ingram et al., 2016, Malkamäki et al., 2017). Nearly three quarters of the world’s industrial forest plantations are composed of Pinus (42%) and Eucalyptus species (26%) (Payn et al., 2015). However, concerns have arisen about the economic and environmental costs of fertilizers and pesticides, productivity losses from pests and diseases and reduced biodiversity in these monospecific production systems (FAO, 1992). Mixed-species plantations have the potential to address these concerns while simultaneously improving nutrient cycling (e.g. Koutika et al., 2017, Liu et al., 2015, Tchichelle et al., 2017), soil fertility (e.g. Montagnini, 2000), biomass production (e.g. Epron et al., 2013, Pretzsch et al., 2013) and carbon sequestration (e.g. Wang et al., 2009, Koutika et al., 2014) as well as providing other benefits through a diversification of products, improved risk management and protection from pests and diseases (Forrester, 2004, Kelty, 2006, Bauhus et al., 2017). Mixed-tree plantations containing N₂-fixing tree species are also thought to provide an additional benefit: a reduced need for nitrogen fertilization thanks to symbiotic N₂ fixation (Forrester et al., 2006a, Piotto, 2008, Bouillet et al., 2013).

However, the success of mixed-tree plantations (i.e. when the mixture is more productive than the monoculture) is highly variable (e.g. Bauhus et al., 2000 for a positive effect, Parrotta, 1999 for a negative effect and DeBell et al., 1987 for no effect). If the interspecific competition in the mixture is more intense than the intra-specific competition in the monoculture, the mixture is likely to be less productive. On the other hand, niche sharing and facilitation, especially when N₂-fixing species are introduced, are expected to promote biomass production in the mixture. However, it is very difficult to predict which kind of interaction will be preponderant and to guarantee the success of the mixture (Forrester et al., 2006a). According to the stress gradient theory (Bertness and Callaway, 1994), positive effects (complementarity) should prevail over negative effects (competition) in a mixture under stressful abiotic conditions. Positive interactions between species (i.e. facilitation and competition reduction) are generally more prevalent in sites with low nutrient availability (Forrester, 2014).

First of all, the design of the mixed-species plantation must be adapted to local conditions to maximize the chances of success. Many options have been illustrated in the literature (Forrester et al., 2006a, Piotto, 2008). Under tropical latitudes, the N₂-fixing species introduced with the economic target species (almost exclusively a eucalypt) most often belong to the Acacia genus, though species from the Leucaena, Casuarina, Albizia or Enterolobium genera are also occasionally used. Under temperate latitudes, N₂-fixing species mostly belong to the Robinia or Alnus genera, and more rarely to the Caragana genus, while the non-fixing species are more diverse: species from the Populus, Salix, Pinus and Pseudotsuga and other genera are used. N₂-fixing species are mainly legumes (Fabaceae Lindl. family) in which N₂ fixation is realized through their symbiosis with bacteria from the genus Rhizobium, except species from the Alnus and Casuarina genera, which form their symbiosis with bacteria from the genus Frankia. The mixing design can take the form of an additive series, where the density of the non-fixing species is kept constant, or a replacement series, where the N₂-fixing trees replace certain non-fixing trees to keep the total planting density constant. Tested proportions used to evaluate experimentally mixture effects range from 11 to 75% of N₂-fixing trees, but a fifty-fifty mixture remains the most widely used option (e.g. Bi and Turvey, 1994).

This study aimed to provide updated and complementary information compared to previously published reviews or meta-analyses, about eucalypt – acacia mixtures (Forrester et al., 2006a), and about forest mixed-species plantations in general (Piotto, 2008, Jactel et al., 2018, Zhang et al., 2012). All these studies suggested that mixed stands were globally more productive than pure ones. Zhang et al. (2012) calculated that mixed-species forests are globally 15% more productive than the average of their component monocultures, and Jactel et al. (2018) estimated that polycultures were 24% more productive than monocultures. However, these two meta-analyses reported that the positive effect of the mixtures was independent of the presence of N₂-fixing species in the mixture.

We carried out a quantitative study compiling the data available in the scientific literature about all kinds of mixed-tree plantations which included N₂-fixing species and undertook a meta-analysis – a set of statistical tools that makes it possible to combine the outcomes of independent studies to evaluate the overall effect of a particular factor and to test the influence of covariates on this effect (Gurevitch and Hedges, 1999). Our main objectives were to calculate a mean effect of mixed-tree plantations on biomass production compared to the monoculture of the non N₂-fixing species from the data reported in the literature. We then sought to evaluate the effects of plantation attributes in terms of (1) climate (temperate vs. tropical), (2) the species used (eucalypt vs. other non N₂-fixing species, and leguminous species vs. other N₂-fixing species), (3) the proportion of N₂-fixing species compared to the non-fixing species (high, low or equal proportions), and (4) the developmental stage for short rotation stands (juvenile or shortly after planting vs. nearing rotation age). Planting density was not tested since only two studies compared this factor. Only replacement series designs were considered in order to hold planting density constant. We chose to compare the mixed-tree plantations to the monocultures of the non N₂-fixing species and not to the monocultures of the N₂-fixing species because we considered that if the N₂-fixing monoculture was more productive than the mixture, the mixture would be useless in economic terms. We tested the following hypotheses: (1) globally, mixed-tree plantations including an N₂-fixing species should be more productive than the monoculture because of the additional nitrogen symbiotically fixed; (2) this better performance of the mixture should be more marked under temperate latitudes where soil nitrogen is generally considered to be less available than in tropical latitudes (Martinelli et al., 1999); (3) a balanced mixing proportion (50/50) would give the best results as this proportion would provide enough N₂-fixing trees to promote biomass production of the non-fixing species and not too many N₂-fixing trees lowering overall stand biomass production; (4) older developmental stages should give better results than juvenile stages since the interactions between species are likely to be limited in very young plantations; it has also been shown that synergistic effects between species are long lasting (Forrester et al., 2004, Zhang et al., 2012).