Effects of leaf litter traits on terrestrial isopod and millipede consumption, assimilation and growth

Nutrient cycling through leaf litter consumption is an essential ecological function performed by macrodetritivorous invertebrates such as isopods and millipedes. Leaf litter consumption rates can vary greatly depending on the environment, consumer identity, and litter traits, but generalizations about the effects of plant traits on macrodetritivore leaf litter consumption, assimilation and growth are not well established and mostly indirectly inferred.
We conducted a systematic search of the global literature and obtained 456 standardized measures from laboratory experiments of relative consumption (RCR), assimilation (RAR) and growth (RGR) rates of terrestrial isopods and millipedes, extracted from 56 different articles. We investigated if commonly measured leaf traits, plant functional groups, prior microbial conditioning of leaves, and climatic conditions affected these rates. We obtained data on commonly measured leaf traits from the TRY global plant trait database, inferred plant functional groups from taxonomic groupings, and obtained climatic data from information reported within articles.
RCR, RAR and RGR varied greatly among macrodetritivore and plant species, but overall, there were no differences between isopods and millipedes. Microbial conditioning of litter greatly increased RCR. Plant functional group was an important predictor of RCR, with eudicot trees and forbs being consumed in greater quantities than magnoliid trees and grasses. Fresh leaf N:P ratio had a positive effect on RAR, and leaf N and C:N ratio had positive and negative effects on RGR respectively, while climatic variables had weak effects on the three rates.
Our work shows that plant traits (both those associated with plant functional groups and commonly measured leaf traits) exert strong effects on resource processing rates by terrestrial macrodetritivores. Further, prior microbial conditioning of leaf litter has a large and globally consistent positive effect on macrodetritivore litter consumption, suggesting that they may consume little, if any, freshly senesced leaf material when microbially conditioned litter is available. Our results suggest that, where extremes of temperature or precipitation do not occur, variables reflective of food quality (leaf traits and microbe conditioning) are more important drivers of macrodetritivore leaf litter consumption than are extrinsic climatic variables.

Methods
We searched in Scopus for all peer-reviewed scientific articles published in English up to October 2023 whose titles, keywords or abstract met the search criteria defined by the Boolean expression: “(isopod* OR millipede OR diplopod* OR ((pill OR sow) AND bug) OR (wood AND (louse OR lice))) AND (feed* OR fed OR consum* OR ingest* OR assimil* OR grow*) AND (rate* OR efficiency)”. This search resulted in 785 articles (see Fig. S1 for the Preferred Reporting Items for Systematic Reviews and Meta-Analyses [PRISMA] diagram). However, some ecological papers that measured the relevant rates but used highly generalizable terminology in titles, abstracts and keywords could not be discovered by this search methodology. We thus added to this collection 17 additional articles identified through our knowledge of the literature, references made by other studies, or identified by data contributors and reviewers. We then manually screened each article and looked for data on RCR, RAR and RGR that were either reported directly by the authors or could be derived from the data provided. Definitions of RCR, RAR, and RGR are described in the online supporting information (supplementary methods). If rates were calculated but the data were not reported in texts or tables, we contacted authors to obtain the raw data, or read the rates off of figures (mostly bar plots) using Fiji ImageJ (Schindelin et al., 2012) if authors could not be contacted or no longer had the raw data. Authors were not contacted if the paper was published more than 30 years ago.

Data were only recorded from studies or treatment levels that approximated physiologically benign (e.g., spring temperatures) and natural current-day field conditions. For example, in studies investigating the effects of elevated CO2 levels (Hattenschwiler et al., 1999; Hattenschwiler & Bretscher, 2001) heavy metal pollution (Calhôa et al., 2006; da Silva Junior et al., 2013; Tourinho et al., 2015; Valant et al., 2012), pesticides (Ribeiro et al., 2001), antibiotics (Zimmer et al., 2001), biodegradable plastics (Wood & Zimmer, 2014), desiccation (Iatrou & Stamou, 1989; Leclercq-Dransart et al., 2019) and salinity (Škarková et al., 2016) on RCR/RAR/RGR, only data from the control groups were utilized. RCRs from the consumption of mixed-species litters or artificial resources were also excluded since plant trait data could not be obtained for these. Only data from terrestrial isopod, millipede and plant species were collected.