Invasive alien shredders clear up invasive alien leaf litter

Abstract Biological invasions have the potential to alter ecosystem processes profoundly, but invaders are rarely found alone. Interactions between different invasive alien species, and their cumulative impact on ecosystem functioning, have led to hypotheses of invasion meltdown whereby effects become additive leading to further ecosystem stress. Invasive riparian plants (e.g., Rhododendron ponticum) deposit leaf litter in freshwaters, which may be unconsumed by indigenous species, potentially affecting habitat heterogeneity and flow of energy to the food web. However, invasive alien decapod crustaceans are effective consumers of leaf litter, and it was hypothesized that they would also consume inputs of invasive riparian leaf litter. This study shows that invasive alien signal crayfish (Pacifastacus leniusculus) and Chinese mitten crab (Eriocheir sinensis) effectively break down different types of leaf litter, including invasive alien R. ponticum, at higher rates than indigenous white‐clawed crayfish. Secondary products were more varied, with more fine particulate organic matter generated for the less palatable alien leaf litter species. Leaf species caused different changes in body mass of decapods but effects were heterogeneous by leaf and decapod: P. leniusculus showed lower mass loss when consuming R. ponticum while E. sinensis lost mass when consuming A. pseudoplatanus. Impacts of riparian invasions on detritus accumulation in freshwaters are thus potentially buffered by invasive alien decapods, illustrating a need for a more detailed consideration of both positive and negative interspecific feedbacks during biological invasions.

Two IAS might consume resources complementarily to result in synergistic resource depletion (Rosewarne et al., 2016). The potential impact of these interactions with multiple IAS on ecosystem functioning has so far been little studied.
These two guilds of IAS (riparian plants, shredders) when considered in isolation have the potential to impact detrital processing in freshwater ecosystems. Leaf litter is converted into smaller fragments by shredding animals, either mechanically through tearing up of the leaf material or digestively through passage through the gut (Doherty-Bone et al., 2018;Montemarano, Kershner, & Leff, 2007). Studies of invasive alien leaf litter have shown both enhanced and reduced decomposition rates by indigenous consumers (freshwater microbes, invertebrates) (Hladyz et al., 2009). Invasive alien omnivores have been found to increase decomposition rates relative to native analogues (Doherty-Bone et al., 2018;Dunoyer, Dijoux, Bollache, & Lagrue, 2014;James, Slater, Vaughan, Young, & Cable, 2015). Yet, the combined effects of invasive riparian plants and invasive alien omnivores on detrital processes have not been investigated.
This study compares the processing of leaf litter from invasive riparian woody plant species by invasive freshwater decapods and contrasts these to equivalent native species. The study focused on two prominent invasive alien plants in the British Isles, Rhodendendron ponticum Linnaeus and Acer pseudoplatanus Linnaeus. Both these species were introduced from mainland Europe and are now widespread in the British Isles (Figure 1a), as well as other regions such as Scandinavia and New Zealand, where they are known to become dominant, shade out other plants and waterways, as well as the leaf litter being of comparable lower quality (Hill & Hulme, 2012;Hladyz et al., 2009Hladyz et al., , 2011Squirrel, 2015). The invasive alien decapod species included in this study are Europe and North America and are rapidly expanding their range (Herborg, Rudnick, Siliang, Lodge, & Macisaac, 2007), likely impacting ecosystems through strong top-down regulation of smaller animals, plants, and detritus (Rosewarne et al., 2016;Rudnick & Resh, 2005). It was hypothesized that (H 1 ) invasive alien decapods would have a higher rate of detrital processing, than native invertebrate shredders as observed by Doherty-Bone et al. (2018); (H 2 ) decomposition rates would be lower for the IAS R. ponticum and A. pseudoplatanus due to higher tannin and cellulose content than the dominant native riparian species Alnus glutinosa (Hladyz et al., 2009). As a consequence, (H 3 ) the assimilation (measured by gain in mass) of invasive alien leaf litter will be higher for invasive alien decapods compared to the native crayfish.

| MATERIAL S AND ME THODS
We compared processing of detritus from riparian plants of a dominant UK riparian native plant (Black Alder, A. glutinosa) to the invasive alien A. pseudoplatanus and R. ponticum by one native (Austropotamobius pallipes) and two invasive (P. leniusculus, E. sinensis) decapods. For the three plants, leaf litter was collected upon abscission around the University of Leeds, dried at 50°C for 24 hr then stored in cool, dry, paper bags in the dark.
While oven drying can cause plant material to leach mass at a faster, exaggerated rate to leaf litter naturally deposited into waterways (Gessner, Chauvet, & Dobson, 1999), this process enables more accurate comparisons of biomass before and after the experiment. Fourteen days prior to the experiment, 1.5 g leaf was placed in mesh bags and microbially conditioned with water from a nearby stream (Doherty-Bone et al., 2018). Austropotamobius pallipes were collected from Wyke Beck, Leeds, UK, P. leniusculus from Fenay Beck, Huddersfield, and E. sinensis collected from the River Thames, Chiswick. Decapods were kept in aged tap water for a minimum of 14 days and unfed for 24 hr hours prior to use in the experiment.
Each microcosm consisted of a 4-L plastic tank with aerated, aged tap water with a PVC pipe shelter for the decapods. A base of nylon mesh (1 mm aperture) allowed organic matter <1 mm to fall through to a separate chamber (Doherty-Bone et al., 2018). These were maintained in an environmental chamber at 14°C on a 16:8 hr light:dark cycle throughout the experiment.

| Experimental design and sample processing
Twelve treatments (10 replicates/treatment, 120 microcosms total) were established containing the conditioned leaf litter from one of the three plant species, and a decapod treatment (native crayfish, invasive crayfish, invasive crab, control without decapod). Subadults of similar sizes were used because they are the most common age class encountered in the field (Doherty-Bone et al., 2018). At the start of the experiment, decapods were weighed following being dabbed dry with a paper towel: A. pallipes -12.87 ± 2.92 g; P. leniusculus -10.33 ± 2.98 g; E. sinensis -11.02 ± 4.28 g. The experiment ran for 7 days. At the end of the experiment, decapods were weighed to determine changes in mass after being dabbed dry with a paper towel.
The remaining water was homogenized and a 50 ml aliquot removed to sample fine particulate organic matter (FPOM -1 mm-0.7 μm) that was then filtered through 0.7 μm GF/F filters. Samples of leaf litter, CPOM, and FPOM were dried separately at 50°C, weighed and ashed at 500°C to estimate ash-free dry mass (AFDM).

| Data analysis
Decomposition rate was calculated as the change in AFDM of leaf litter per day, final AFDM estimated based on the baseline AFDM estimated from a separate batch of conditioned leaf litter of each species (Benfield, 2006). As mass of a decapod can influence consumption rates of leaf litter, as well as to enable estimation of impact by mass of animal, rates of decomposition and production of residual materials were divided by decapod mass to show mass-specific performance (Doherty-Bone et al., 2018). This enabled measurement of both the actual impact (decomposition rate, etc.) and the quantitative functional trait of detritivory performance (decomposition rate per mass of animal). Generalized linear models (formula: glm, R v.3.1.0.; R Development Core Team, 2014) were used to compare decomposition rate, CPOM and FPOM production, and change in decapod mass against leaf and decapod species treatment, with decapod mass included as a covariate. The models were fitted based on the distribution of the data (i.e., family = Gaussian; Poisson) identified using maximum likelihood estimates (library: MASS, formula: fitdistr). Model fit of the glm was compared using Akaike information criterion (AIC) (Burnham & Anderson, 2004) and Tukey post hoc tests applied to establish pairwise differences  Source: National Biodiversity Network Atlas www.nbnatlas.org Accessed 01 April 2018 (see Supporting information Appendix S1 for details of distributional data sources)

| RE SULTS
Both leaf species and decapod species had a significant effect on leaf decomposition rate and production of FPOM, the interaction between leaf and decapod species showing the highest performing models (ΔAIC = 0) ( Table 1). Production of smaller fragments of CPOM was affected by decapod species but not by leaf species (Table 1). Decomposition rates were similar between the invasive alien A. pseudoplatanus and the native A. glutinosa but lowest for R. ponticum (Figure 2a, Supporting information Table S1). There were significant interactions between leaf species, decapod species, and decapod mass for decomposition rate (albeit with a low likelihood based on a high ΔAIC, Table 1), with higher decomposition rates for the invasive alien P. leniusculus and E. sinensis than for the native A. pallipes (Figure 2a). Notably, when compared with controls and the native A. pallipes treatment, R. ponticum leaf litter decomposed faster in the invasive P. leniusculus and E. sinensis treatments.
Invasive alien decapods produced more residual CPOM (a likely by-product of uneaten leaf litter that is difficult to handle and be consumed by larger animals) than A. pallipes and controls, but did not differ significantly across leaf species (Figure 2b). FPOM was significantly influenced by the interaction between decapod and leaf species (ΔAIC = 0, Table 1). For all leaf species, significantly more FPOM was produced by invasive alien decapod species than by A. pallipes and controls (Figure 2c). A. glutinosa leaves resulted in significantly less FPOM production compared to A. pseudoplatanus and R. ponticum (Figure 2c). Mass-specific variables such as decomposition rate, CPOM, and FPOM production expressed as change in mass per decapod body mass, all showed significant relationships to decapod and leaf species. These showed the same relationships as the actual decomposition rates, CPOM, and FPOM production: invasive alien decapods showing higher mean values (Figure 2). Efficiency of leaf litter decomposition showed the best fitting model for leaf and decapod species, as did CPOM and FPOM production efficiency ( Table 2).
The change in decapod mass at the end of the experiment did not differ overall between decapod species, but was significantly affected by leaf species treatment and the interaction between decapod and leaf species (Table 1). The interaction between leaf and decapod species provided the best performing model (ΔAIC = 0).
Post hoc tests showed heterogeneous mass change in relation to leaf type among decapod species. There was significantly greater mass gain by P. leniusculus through the consumption of A. glutinosa and A. pseudoplatanus leaves and to a lesser degree R. ponticum

| D ISCUSS I ON
This study provides new insights into the impacts of invasive decapods on invasive alien leaf litter and vice versa. These findings are significant because all four of these IAS are expanding their range in the same regions, often co-occurring, yet little is known about the consequences of these multi-invasions. Invasive alien decapods (P. leniusculus and E. sinensis) processed all species of leaf litter at a faster rate than did the native analogue (A. pallipes). The invasive decapods showed higher decomposition rates, converting leaf mass into smaller fragments, and most notably into FPOM. They also processed more of the invasive R. ponticum than did the native decapod.
FPOM production was significantly lower for the native leaf This study provides a novel example of one IAS providing biotic resistance to the effects of another on a key ecosystem process. Rhododendron ponticum causes the build-up of leaf litter in the benthos that is not nutritionally available to microbes or invertebrates, causing shift in benthic food webs and habitat structure in the benthos (Hladyz et al., 2011). While A. pseudoplatanus also provides a substantial amount of leaf litter, this is more palatable to aquatic organisms, thus R. ponticum by comparison is substantially more impactful (Hladyz et al., 2009). Although the invasive decapods processed more R. ponticum, there was negligible growth of decapods consuming R. ponticum, suggesting there would be no selective advantage to consuming this material when more nutritious resources are available. In contrast, A. pseudoplatanus litter has similar impacts (decomposition rates, N:P, fungal biomass, invertebrate colonization) to other native species, particularly A. glutinosa (Hladyz et al., 2009). Testing this interaction in a more realistic mesocosm or in situ cage experiment with multiple available resources would help resolve whether this holds true for real-world invasion scenarios.
Biological resistance to the establishment of newly colonizing species, including IAS, has been a commonly measured variable of ecosystem functioning (Fargione & Tilman, 2005). Native species providing resistance to the impacts of IAS have not been explicitly studied. This is pertinent for native species with high risk of extinction and low ecological redundancy, but with non-native species replacing their role, as for freshwater decapods in most regions (Lodge et al., 2012).  (Hladyz et al., 2011;Maclean et al., 2017) and could also be impacted by depletion of invertebrate prey if they are excluded by the low nutritional value of the leaf litter.

Austropotamobius pallipes processed
Potential management applications of this study include prioritizing interventions in waterways selected for staggered removals of first invasive alien riparian plants, then decapods. This would result in processing of the invasive alien leaf litter by the decapods as the source of the litter is removed. As both P. leniusculus and E. sinensis were also found to maintain, even increase body condition through consuming this invasive alien leaf litter, removal of this riparian resource could potentially, temporarily slow population growth of the invasive alien decapods. Similar recommendations have been made for removal of invasive alien predators after their more impactful prey (Miyake & Miyashita, 2011). However, given the severe impacts and difficulty of control of these invasive alien decapods, rapid response to early introductions of invasive alien decapods should be higher priority.
Invasive alien decapods were shown to remove invasive alien leaf litter, reducing the negative impact of that leaf litter through provision of its otherwise unavailable resources to the remainder of the food web. This provides a clear example of IAS introducing complementarity to biodiversity-ecosystem relationships: Here addition of IAS decapods increased transformation of resources provided by native and alien plants, whereby addition of a species increases a measured ecosystem process such as transformation and capture of a resource (Cardinale, Palmer, & Collins, 2002). The observed interaction of multiple IAS led to a combined, potentially buffering effect on ecosystem functioning.

ACK N OWLED G M ENTS
TMD-B was supported by a John Henry Garner Scholarship and the University of Leeds, and AMD by NERC NE/G015201/1. A. pallipes were collected under Natural England license #2015-17131-SCI-SCI.
The authors thank the laboratory staff of the University of Leeds School of Geography for technical assistance and The Port of London Authority for consent to collect animals. No ethical permission was required for this work.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interests.

AUTH O R CO NTR I B UTI O N S
TDB, AMD, and LEB conceived the study. TDB and JB carried out the experiments. TDB, AMB, JB, and LEB analyzed the data. TDB, AMD, JB, and LEB wrote the manuscript.