Laboratory Leaching Experiment and Field Data Regressions

SUPPORTING INFORMATION

Leaf collection area

Senescent and freshly fallen leaves of Nectandra megapotamica (Spreng.) Mez were collected from a forested area (0.27 km2 of area; coordinates in decimal degrees of -27.6498, -52.2702; Fig. S1) inside the campus of URI (Universidade Regional Integrada do Alto Uruguai e das Missões) in the municipally of Erechim, Brazil. Leaves were collected from ca. 5 adult trees (> 10 m tall).

Fig. S1. Location of the collection area of Nectandra megapotamica leaves used in the experiment.

Laboratory leaching experiment and field data regressions

To evaluate the extent of leaching phase we added ca. 0.75g of air-dried leaves (Nectandra megapotamica; which corresponded of ca. 2-3 entire leaves) in 250 ml recipients with distillated water and maintained in ambient temperature (air temperature ranged between 20-24ºC). We used four replicates for each sampling time (1, 2, 24 and 48 hours) and measured the remaining mass from leaves (after 72h at 70ºC) and the dissolved organic carbon from the water (analyzed in TOC-VCSH, Shimadzu®).

The time of removal litter bags after the leaching phase was set to approximate 25% of initial mass loss based on previous experiments in the region (Tonin et al. (2014), Tonello et al., 2016), which resulted in 15 days. The third removal of litter bags (set to approximate 50% of initial mass loss) was determined based on mass loss regression with data from the first and second removal of litter bags. We selected only data of the two fasted reaches (to have sufficient mass for further analysis) to predict the number of the days required to reach 50% of initial mass loss, based on linear models.

Fig. S2. (a) Leaf mass remaining and water dissolved organic carbon during leaching over 48h experiment; and, (b) linear regressions of mass remaining of leaves incubated in two reaches (SR1 and SR2) at two sampling times (3 and 15 days) and the predicted number of days to reach 50% of leaf mass loss.

Fig. S3. Raw decomposition rates (total, microbial and invertebrate-driven) for each block (one litter bag of each sampling time within a reach) and the estimated variability across the three spatial scales: watershed (W1 – W4), segment (S1 – S8) and reach (R1 – R8). Residual variation represents the variation among litter bags within reaches.

Fig. S4. Comparison of the variability of litter decomposition rates of each stream (estimated through coefficient of variation of k total and k microbial) of the present study (black circles) with other studies in different regions and biomes. Red and orange circles (subtropical streams) = Tonello et al. (2016) and Tonin et al. (2014), respectively; green circles (temperate streams) = Tiegs, Akinwole & Gessner (2009); and, blue circles (tropical streams) = Boyero et al. (2015).

Supplementary References

Boyero, L., Pearson, R.G., Gessner, M.O., Dudgeon, D., Ramírez, A., Yule, C.M., et al. (2015) Leaf-litter breakdown in tropical streams: is variability the norm? Freshwater Science, 34, 759-769.

Tiegs, S.D., Akinwole, P.O. & Gessner, M.O. (2009) Litter decomposition across multiple spatial scales in stream networks. Oecologia, 161, 343-351.

Tonello, G., Naziloski, L.A., Tonin, A.M., Restello, R.M. & Hepp, L.U. (2016) Effect of Phylloicus on leaf breakdown in a subtropical stream. Limnetica, 35, 243-252.

Tonin, A.M., Hepp, L.U., Restello, R.M. & Gonçalves Jr, J.F. (2014) Understanding of colonization and breakdown of leaves by invertebrates in a tropical stream is enhanced by using biomass as well as count data. Hydrobiologia, 740, 79-88.