The impact of invertebrate decomposers on plants and soil.

Soil invertebrates make significant contributions to the recycling of dead plant material across the globe. However, studies focussed on the consequences of decomposition for plant communities largely ignore soil fauna across all ecosystems, because microbes are often considered the primary agents of decay. Here, we explore the role of invertebrates as not simply facilitators of microbial decomposition, but as true decomposers, able to break down dead organic matter with their own endogenic enzymes, with direct and indirect impacts on the soil environment and plants. We recommend a holistic view of decomposition, highlighting how invertebrates and microbes act in synergy to degrade organic matter, providing ecological services that underpin plant growth and survival.

Along with fire, decomposition is the primary process by which the stores of nutrients and carbon captured by plants are recycled through the biosphere, converting dead organic material to simpler forms, which are available again for plants (Bishop et al., 2020;Pausas & Bond, 2020).
While microbes are considered by many as the primary agents of plant decomposition in terrestrial ecosystems (e.g. Crowther et al., 2019;Lustenhouwer et al., 2020;Pausas & Bond, 2020), here, we synthesise recent advances in understanding of the roles that soil fauna play in biomass degradation and biogeochemical cycling, and the consequences of these processes for plants. In doing so, we build on previous studies highlighting the importance of invertebrates in terrestrial recycling pathways (Swift et al., 1979;Wall et al., 2008;García-Palacios et al., 2013;Briones, 2018) and emphasise the key role of invertebrates in decomposition and, crucially, that their contribution to plant growth, nutrition and survival should not be overlooked.
Decomposition has, traditionally, been thought to be hierarchically controlled by climate and litter quality with soil organisms exerting a comparatively weaker influence on decay rates globally (Hättenschwiler et al., 2005;Cornwell et al., 2008;Makkonen et al., 2012). However, this longheld view has recently been challenged by work showing that microbial communities and microclimate can exert an equally important influence on decay rates at local scales than those exerted by latitudinal gradients in climate (Bradford et al., 2016;Bradford et al., 2017). These findings have implications for plants, because sessile organisms are reliant on decomposition processes that drive the availability of soil nutrients in their immediate surroundings. Yet, the relative contribution of soil fauna to the factors controlling fine-scale heterogeneity in decay rates, and therefore the availability and accessibility of soil nutrients to plants, remains to be quantified despite invertebrate decomposers contributing significantly to the breakdown of dead organic matter across the globe. García-Palacios et al. (2013), for example, demonstrated that the exclusion of soil fauna from leaf litter decomposition bags reduced litter mass loss by an average of 35% across seven biomes. The factors mediating deadwood decay and other substrates (e.g. dung) are less well understood globally, with tropical deadwood decomposition studies, for example, representing just 14% of the published decomposition literature (Harmon et al., 2020).
However, in tropical and subtropical systems, evidence is mounting that invertebrate decomposers (termites, in particular) are instrumental for the decomposition of coarse woody material, where they have been shown to be equally, if not more, important than free living microbes for deadwood mass loss (Griffiths et al., 2019;Griffiths et al., 2021;Guo et al., 2021).
The majority of terrestrial above-ground plant biomass is concentrated in tropical ecosystems (Crowther et al., 2019), which means that the decomposition of the majority of dead plant material Accepted Article occurs in the tropics. Yet, the temperate bias in decomposition studies (Guerra et al., 2020) coupled with the dogma that invertebrates are not 'true' decomposers but are instead only facilitators of microbial decay (Crowther et al., 2019;Jones et al., 2020;Lustenhouwer et al., 2020) means that we lack an in-depth understanding of the biotic agents controlling the breakdown of the bulk of plant material on the planet. Furthermore, studies focussed on the consequences of decomposition for plant-soil interactions and associated outcomes for plant communities continue to overlook invertebrate contributions across all ecosystems (e.g. Baskaran et al., 2017). This inaccurate simplification hinders our ability to predict and mitigate the consequences of anthropogenic changes to the below-ground communities that underpin soil biogeochemical cycling and plant growth, nutrition and survival.
Here, we seek to redress this microbe-biased understanding of decomposition by: 1) summarising recent literature demonstrating that many invertebrates are true decomposers able to chemically break down dead plant material; 2) exploring the direct and indirect ways that invertebrate decomposers influence the soil environment and therefore plant growth, nutrition and survival; and 3) highlighting future research directions to deepen our understanding of invertebrate contributions to plant-soil interactions. (See Box 1 for glossary of terms used in this Tansley insight.)

II. Decomposition: chemical breakdown by microbes and invertebrates
Decomposition depends on enzymes that can degrade lignocellulose (i.e. cellulases, hemicellulases and lignases; Cragg et al., 2015). The idea that invertebrates are not directly involved in the catabolism of dead plant material was based on the notion that animal decomposers do not produce endogenous cellulases and appeared to have very low assimilation efficiencies ( Van der Drift, 1951). Yet, in 1998, the first insect cellulase-encoding gene was found in a termite species -Reticulitermes speratus (Watanabe et al., 1998). Since then, cellulaseencoding genes have been isolated in various insects and other invertebrates (Chang & Lai, 2018). Crucially, although many invertebrates rely on a partnership with gut microbes for the breakdown of dead plant material (Cragg et al., 2015) and white-rot fungi are the main organisms capable of degrading lignin (Ayuso-Fernández et al., 2019), a growing number of animals have been shown to degrade organic material independently of microbial partners (Scrivener et al., 1989;Shelomi et al., 2020).
The idea that animal decomposers are capable of only poor assimilation efficiencies was most recently challenged by David (2014) who demonstrated that macroarthropods, such as Accepted Article millipedes and woodlice, are able to digest more than 50% of the dry plant matter they consume. This study, by using 14 C-labelled leaf substrate, confirmed that at least 38% of this material was assimilated or respired by the millipede Glomeris marginata. Furthermore, the digestion of leaf litter by macroarthropods (millipedes, woodlice and snails) has been shown to chemically degrade plant material (Joly et al., 2020). This feeding activity converts huge quantities of leaf litter into faeces, which, when compared with unconsumed leaf litter, has lower C:N ratios and tannin concentrations; higher dissolved organic carbon and total dissolved nitrogen concentrations (Joly et al., 2020); can have higher overall surface area (Joly et al., 2018), and harbour higher microbial biomass and distinct microbial communities (David, 2014). Crucially, these physical and chemical changes accelerate C cycling by between 38 and 50% (Joly et al., 2020) and have been shown to result in a switch from net N immobilisation in unprocessed leaf litter to net N release in the faeces of the millipede Glomeris marginata (Joly et al., 2018).
Therefore, growing evidence suggests that invertebrates do chemically break down dead plant material and may be more important in terrestrial degradation pathways than previously thought.
Building on these advances, Scharf (2015) promoted the idea of the digestome, which describes the combined enzymes produced both by microbes and invertebrates, as a key factor in plant decomposition ( Fig. 1).

III. Decomposition and soil processing by invertebrates and the consequences for plants
Ecologically important decomposer invertebrates include earthworms, termites, woodlice, snails, millipedes, beetles (especially their larvae) and mesofauna such as collembola. There is a wealth of experimental evidence demonstrating these invertebrate decomposers contribute significantly to the mass loss of dead plant matter across the globe (e.g. García-Palacios et al., 2013;Fujii et al., 2018;Griffiths et al., 2019;Yang & Li, 2020;Griffiths et al., 2021;Guo et al., 2021). However, few studies look beyond the effects of soil fauna on decay rates to the consequences for plants, and the majority of those investigations have been carried out in temperate regions or laboratories (e.g. Setälä & Huhta, 1991;Bardgett & Chan, 1999;Eisenhauer et al., 2018;Winck et al., 2020). Invertebrate decomposers can affect plants via trophic effects, which influence soil nutrient mineralisation as a result of enzymatic degradation within the gut (digestive effects on the nutrient status of the decomposing substrate and/or soil e.g. Joly et al., 2020;Joly et al. 2018;Winck et al., 2020); and/or via non-trophic effects resulting from movement and nest building, which alters soil structure (soil particle distribution, aeration, soil moisture Consequently, our knowledge of the mechanisms by which invertebrate decomposers drive plantsoil interactions is patchy and incomplete.

Trophic effects of invertebrate-mediated decomposition on soil nutrients and plants
Experimental work has demonstrated that the direct positive effect of invertebrate decomposers on soil nutrient availability can translate into benefits for plant growth and nutrition (Pathway 2, ). The magnitude of these effects, in agroecosystems, was found to be dependent on the Accepted Article presence of crop residue, earthworm density and the type and rate of fertilization, indicating that earthworms may stimulate plant growth predominantly through releasing nitrogen locked away in residues and soil organic matter (Li et al., 2020). Uncertainties remain in disentangling the relative contributions of direct invertebrate-mediated changes in soil nutrient cycling (e.g. Joly et al., 2020) from those driven by shifts in microbial communities (Des Marteaux et al., 2020;Bray et al., 2019). However, what is becoming clearer is the vital role that invertebrate-microbial partnerships (rather than either group in isolation) play in driving decomposition and the importance of these interdependent processes for plants.

Non-trophic effects of invertebrate movement and nest building
The indirect effects of invertebrate decomposers, via their movement through the soil and

IV. Outlook
Human activity, including climate change, land-use change, biological invasions and nutrient deposition are leading to changes in soil fauna diversity, abundance and distributions (Geisen et al., 2019). Shifts in any part of complex soil food-webs could affect the ecosystem processes carried out by invertebrate decomposers, with cascading consequences for plants.
Given that direct evidence for which soil invertebrate activities affect plants is so incomplete, we have limited capacity to predict the consequences of loss in below-ground biodiversity for ecosystem functioning. However, in light of the evidence we present here, which shows that soil Consequently, we have limited understanding of how these processes shift with environmental change. A more interdisciplinary approach is necessary to understand the complex partnerships between soil microbes and invertebrates, and the consequences of these for plant nutrition, growth and survival in our era of rapid global change.

Accepted Article
This article is protected by copyright. All rights reserved

Accepted Article
This article is protected by copyright. All rights reserved Decomposition: The process of breaking down dead organic (in this case) plant material into smaller fragments and/or molecules (Lehmann & Kleber, 2015) either by abiotic agents (e.g. photodegradation) or decomposer organisms through catabolism by microbial enzymes (most importantly cellulases) and/or invertebrate endogenous enzymes in concert with microbial symbionts.

Digestome:
The whole set of enzymes found in an invertebrate gut: the endogenous enzymes, mostly produced in the mid gut and/or salivary glands, and the exogenous enzymes produced by symbionts in the hind gut (or farmed externally, in the case of fungus-growing termites). The combined digestome is the agent of invertebrate decomposition.
Endogenous cellulase: Cellulases produced by organisms within their own tissues, as opposed to exogenous cellulases, which are produced by symbionts.

Invertebrate functional classifications:
There are many functional classifications of soil fauna.
Here we concentrate on decomposers, defined as any group that ingests dead plant material leading to changes in physiochemical composition, contributing to a reduction in the size of fragments and/or molecules and facilitating the interaction with mineral surfaces in soil aggregates, therefore stabilising and protecting organic material from further degradation (Fig. 1).
Animals that feed on animal carcasses are also decomposers and make nutrients available to plants.

Mutualistic symbionts:
Organisms living with another organism each of which conferring benefits to the other. Here, we predominantly focus on the gut biota of invertebrates.
Plant chemistry and decomposability: Plants have interior cell chemicals similar to other organisms (i.e. sugars, fats and proteins, most of which are easy to metabolise by nearly all organisms). However, all plants have cell walls made of cellulose, a complex polymer which is a linear chain of several hundred to many thousands of β linked D-glucose units (Cragg et al., 2015). It is much harder to metabolise cellulose and far fewer organisms have the enzymes that can depolymerise it. The cellulose fibres are, in turn, linked by hemicellulose chemical cross linkages. In addition, many plants, especially woody plants, have a substantial amount of lignina cross-linked phenolic polymer, which is catabolised by relatively few organisms (all are fungi or bacteria, most efficiently by white rot fungi).
Soil fertility refers to the availability of soil nutrients and is often used in agricultural contexts (e.g. how well a particular crop grows in a soil).

Box 2 Future directions
Understanding which invertebrates have endogenous cellulases. While our knowledge is improving, there remain large gaps in our understanding of which invertebrate taxa can be considered 'true' decomposers (with endogenous cellulases), how ubiquitous they are, and how these properties work in partnership with symbiotic microorganisms (i.e. the digestome).
Field-based manipulations of invertebrate decomposers replicated at high spatial resolution are needed to assess their effects on vegetation decay rates with contrasting traits across a range of environments. This will fill major gaps in understanding of invertebrate contribution to the hierarchy of controls of decomposition (Hättenschwiler et al., 2005;Cornwell et al., 2008;Makkonen et al., 2012;Bradford et al., 2016;Bradford et al., 2017) and enable quantification of how small-scale heterogeneity in soil fauna, microbes and microclimate shape decay rates and biogeochemical cycling. Furthermore, rather than focussing on litter quality within the same class of substrate, manipulations using different substrate types (e.g. wood, litter, dung), will allow a more complete understanding of the role that different agents of decomposition play for different substrates. These data are essential to accurately parameterise Earth System Models incorporating real-world heterogeneity in biotic communities and ecological processes.

Accepted Article
This article is protected by copyright. All rights reserved beyond assessing litter mass loss and also assess the concomitant impacts on soil nutrient status, physical structure, microbial communities and consequences for plant growth and survival.
Fruitful avenues include using isotope tracers to follow the fate and assimilation rate of carbon and nitrogen from dead organic matter through the soil food web (e.g. David, 2014;Chomel et al., 2019) and the consequences of this for plants.

Contributions of invertebrates to biogeochemical cycles and incorporation into Earth
System Models. Reductions in decomposer diversity (both microbes and soil invertebrates) tend to have negative effects on rates of decomposition (Srivastava et al., 2009;Handa et al., 2014).
Despite this, and although incorporation of soil fauna has been shown to fundamentally affect the predictive outcome of soil organic matter models (explored in Filser et al., 2016), invertebrate contribution to decomposition is not currently incorporated into Earth System modelling.
Therefore, quantifying the contribution of animal decomposers to nutrient and carbon cycling and incorporation of these data into Earth System Models is a research priority that represents a major challenge. Implementation requires a synthesis of emerging technologies and traditional field experimentation to combine: 1) remote sensing techniques, which are increasingly able to detect, map and predict land-scape scale variability in vegetation as well as microclimatic conditions at the land-air interface (Zellweger et al., 2019); 2) detailed field experiments that partition the role of invertebrates in biomass degradation, biogeochemical cycling, carbon flux and plant growth and survival. Field experiments should ideally be carried out with a high levels of spatial replication within focal ecosystems to capture biogeochemical responses to fine-scale environmental heterogeneity; 3) high-throughput DNA sequencing to provide high resolution taxonomic information on which of the soil organisms are driving soil processes and plant responses observed in field manipulations; and 4) Earth System modelling, able to process and integrate the huge volumes of data generated from these disparate research fields. This interdisciplinary approach will allow us to model and predict the consequences of interdependent changes in biotic communities and abiotic conditions for terrestrial degradation pathways and the biogeochemical cycles that underpin vegetation communities and regulate global climate.
Distribution of decomposers in a changing world. Anthropogenic driven distribution shifts in soil fauna are likely to change biogeochemical processing in ecosystems. Advances in molecular techniques will make mapping the relationships between changes in decomposer distributions, decomposition, carbon and nutrient cycling and plant responses more easily quantifiable, and should be a research priority. Field-based experiments that simulate anthropogenic impacts such as drought are necessary to underpin modelling efforts to understand how changes to Accepted Article decomposer fauna will shift under climate change. To date there has been a focus on agricultural ecosystems and earthworms in temperate zones, but more work on other soil invertebrates in other ecosystems would provide insight into their roles in ecosystem function and how this will be altered by environmental change. Figure 1 Tansley Insight 35032 Tansley Insight 35032

Consequences for soil properties
Trophic effects Non-trophic effects Unresolved effects 1 2 3 2 3