Dias ATC1*, Krab EJ1, Mariën J1, Zimmer M2, Cornelissen JHC1,Ellers J1, Wardle DA3 and Berg MP1. Traits underpinning desiccation resistance explain distribution patterns of terrestrial isopods. Oecologia.
1 Department of Ecological Science, Faculty of Earth and Life Sciences, VU University Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, The Netherlands
2 FB Organismische Biologie: Ökologie, Biodiversität & Evolution der Tiere, Paris-Lodron-Universität, Hellbrunner Str. 34, 5020 Salzburg, Austria
3 Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, SE 90183 Umeå, Sweden
*Correspondence author: /
Online Resource 1:Collection sites for each species in 2010 and 2005. Trait values (mean ± SD) are shown: ventral surface area (mm2), water loss rate (mg mg-1h-1) when exposed to 85% and 36% relative humidity (WLR_85%HR and WLR_36%HR), fatal water loss (FWL, proportion of water lost at the moment of death) and desiccation resistance (DR, survival time in hours).
2010 / 2005Species / Site / Habitat / Surface area / WLR_85%RH / FWL / DR / Site / Habitat / Surface area / WLR_36%RH
A. dentiger / Maastricht (50.760°, 5.925°) / forest / 7.3 ± 0.5 / 0.093 ± 0.010 / 0.55 ± 0.04 / 5.5 ± 0.5 / - / - / - / -
A. album / Borselle (51.423°, 3.704°) / beach / 8.0 ± 1.8 / 0.014 ± 0.003 / 0.56 ± 0.05 / 39.4 ± 7.6 / Borselle (51.423°, 3.704°) / beach / 13.1 ± 1.2 / 0.05 ± 0.01
A. opacum / Maastricht (50.760°, 5.925°) / forest / 22.3 ± 4.6 / 0.009 ± 0.001 / 0.49 ± 0.06 / 57.6 ± 10.1 / Maastricht (50.760°, 5.925°) / forest / 28.5 ± 3.6 / 0.02 ± 0.00
A. pictum / Maastricht (50.760°, 5.925°) / forest / 14.0 ± 2.7 / 0.008 ± 0.001 / 0.55 ± 0.04 / 74.7 ± 13.8 / Maastricht (50.760°, 5.925°) / forest / 13.5 ± 2.4 / 0.04 ± 0.01
A. pulchellum / Vasse (52.451°, 6.828°) / forest / 9.3 ± 1.2 / 0.015 ± 0.008 / 0.54 ± 0.10 / 45.9 ± 26.1 / - / - / - / -
A. vulgare / Maastricht (50.760°, 5.925°) / forest / 30.9 ± 13.4 / 0.006 ± 0.001 / 0.43 ± 0.11 / 73.6 ± 18.5 / Nijmegen (51.866°, 5.982°) / grassland / 27.6 ± 4.7 / 0.03 ± 0.00
E. caelatum / Goes (51.534°, 3.917°) / dike / 19.0 ± 2.0 / 0.020 ± 0.002 / 0.40 ± 0.14 / 19.7 ± 8.0 / Goes (51.534°, 3.917°) / dike / 20.8 ± 2.9 / 0.05 ± 0.01
H. danicus / Nijmegen (51.866°, 5.982°) / forest / 1.9 ± 0.5 / 0.137 ± 0.027 / 0.38 ± 0.08 / 2.8 ± 0.4 / Amsterdam (52.324°, 4.888°) / grassland / 1.6 ± 0.5 / 0.35 ± 0.05
H. mengii / - / - / - / - / - / - / Heerlen (50.868°, 5.969°) / grassland / 1.1 ± 0.3 / 0.37 ± 0.06
H. riparius / - / - / - / - / - / - / Nijmegen (51.866°, 5.982°) / forest / 5.1 ± 1.8 / 0.32 ± 0.06
L. oceanica / Borselle (51.423°, 3.704°) / salt marsh / 64.3 ± 13.9 / 0.012 ± 0.002 / 0.53 ± 0.07 / 46.8 ± 15.7 / Texel (52.998°, 4.746°) / salt marsh / 49.1 ± 9.5 / 0.06 ± 0.01
L. hypnorum / Ootmarsum (52.378°, 6.914°) / forest / 10.7 ± 1.6 / 0.097 ± 0.028 / 0.45 ± 0.07 / 4.5 ± 0.9 / Valkenburg (50.878°, 5.813°) / forest / 12.0 ± 3.1 / 0.16 ± 0.04
M. leydigii / Rustenburg (52.630°, 4.870°) / grassland / 1.4 ± 0.4 / 0.192 ± 0.043 / 0.44 ± 0.08 / 2.3 ± 0.4 / Schoorldam (52.701°, 4.706°) / grassland / 1.8 ± 0.5 / 0.35 ± 0.03
M. patiencei / Hoedekensdijk (51.417°, 3.906°) / dike / 1.8 ± 0.4 / 0.171 ± 0.023 / 0.45 ± 0.09 / 2.7 ± 0.7 / Hoedekensdijk (51.417°, 3.906°) / dike / 1.7 ± 0.3 / 0.42 ± 0.07
O. asellus / Maastricht (50.760°, 5.925°) / forest / 31.9 ± 7.4 / 0.021 ± 0.003 / 0.46 ± 0.09 / 22.1 ± 7.3 / Maastricht (50.760°, 5.925°) / forest / 42.4 ± 3.9 / 0.05 ± 0.00
P. muscorum / Nijmegen (51.866°, 5.982°) / forest / 11.5 ± 2.2 / 0.015 ± 0.003 / 0.45 ± 0.14 / 31.8 ± 12.1 / Maastricht (50.760°, 5.925°) / forest / 17.5 ± 2.5 / 0.05 ± 0.01
P. hoffmannseggi / Maastricht (50.760°, 5.925°) / grassland / 2.8 ± 0.7 / 0.124 ± 0.028 / 0.46 ± 0.04 / 3.6 ± 0.9 / Schoorldam (52.701°, 4.706°) / grassland / 3.5 ± 0.6 / 0.28 ± 0.04
P. scaber / Amsterdam (52.324°, 4.888°) / forest / 28.5 ± 4.6 / 0.009 ± 0.002 / 0.44 ± 0.08 / 51.3 ± 14.8 / Maastricht (50.760°, 5.925°) / forest / 23.9 ± 5.0 / 0.04 ± 0.01
P. spinicornis / Millingen (51.874°, 6.026°) / dike / 28.7 ± 4.1 / 0.007 ± 0.005 / 0.49 ± 0.06 / 83.5 ± 29.5 / Den Oever (53.017°, 5.193°) / dike / 35.5 ± 19.7 / 0.02 ± 0.02
P. pruinosus / - / - / - / - / - / - / Wageningen (51.984°, 5.649°) / compost heap / 21.3 ± 4.5 / 0.03 ± 0.01
P. conspersum / Valkenburg (50.878°, 5813°) / forest / 10.7 ± 2.2 / 0.024 ± 0.003 / 0.45 ± 0.07 / 18.1 ± 4.4 / Valkenburg (50.878°, 5.813°) / forest / 12.1 ± 1.4 / 0.08 ± 0.01
T. rahtkii / Rustenburg (52.630°, 4.870°) / grassland / 23.8 ± 5.9 / 0.020 ± 0.004 / 0.45 ± 0.09 / 22.6 ± 6.6 / Nijmegen (51.866°, 5.982°) / grassland / 21.2 ± 5.2 / 0.07 ± 0.01
T. albidus / Rustenburg (52.630°, 4.870°) / grassland / 1.6 ± 0.5 / 0.238 ± 0.038 / 0.53 ± 0.09 / 2.2 ± 0.4 / Rustenburg (52.630°, 4.870°) / grassland / 2.8 ± 0.6 / 0.36 ± 0.05
T. helveticus / - / - / - / - / - / - / Nijmegen (51.866°, 5.982°) / grassland / 1.3 ± 0.1 / 0.36 ± 0.04
T. sarsi / Rustenburg (52.630°, 4.870°) / grassland / 1.6 ± 0.2 / 0.175 ± 0.045 / 0.49 ± 0.07 / 2.6 ± 0.4 / Rustenburg (52.630°, 4.870°) / grassland / 1.8 ± 0.3 / 0.42 ± 0.05
T. pusillus / Valkenburg (50.878°, 5.813°) / grassland / 2.1 ± 0.5 / 0.176 ± 0.044 / 0.44 ± 0.11 / 2.4 ± 0.6 / Nijmegen (51.866°, 5.982°) / grassland / 2.7 ± 0.5 / 0.39 ± 0.04
Online Resource 2: Schematic overview of the experimental setup to measure desiccation resistance. I) Acclimation procedure to replenish possible water deficits. A closed glass box with water saturated plaster element was used, which can be filled via a pipe (A). Animals were kept individually on small cylinders (B) with the bottom made of moist plaster and closed with gauze on top. These cylinders were placed in tight contact with the layer of moist sand that is on top of the plaster element. II) Exposure of animals to dry conditions. A falcon tube (C) was filled with 20mL of glycerol solution to determine relative humidity in the headspace. The animal to be measured was placed in an open-top chamber (D) with the bottom made of plastic mesh (0.5mm wide). The open-top chamber was placed on top of a platform made of plastic mesh (2mm wide), allowing gas exchange between the glycerol solution and the headspace. To weigh the animal repeatedly in order to calculate the water loss rate, the whole open-top chamber was removed, weighed and immediately returned to the falcon tube. In this way it was possible to record changes in the animal mass quickly
and with minimal disturbance.
Online Resource3: Graphs showing examplesofwater loss with time at 85%RH and 15°Cfor individuals of two species. Longer gaps between data points represent overnight periods when no measurements were performed. (a) Calculation of water loss rate, desiccation resistance (survival time, observed visually) and fatal water loss. Initial water content (Iwc) and final water content (Fwc) are shown. Water loss rate was calculated as the slope of the regression between water mass and time divided by the initial water content. Fatal water loss was calculated as Iwc-Fwc/Iwc. (b) Graph showing that water loss remains constant, possibly even after the animal has been dead for part of the night. The right-most data point (highlighted) indicates when the animal was found dead.
Online Resource 4: Methods and detailed results of Phylogenetic analyses.
Methods
We used the sequences of 18S ribosomal RNA to construct a phylogenetic hypothesis for the terrestrial isopods that occur in the Netherlands. The 18S rRNA sequences were retrieved from NCBI for 14 species. In addition, we sequenced the 18S rRNA of 16 novel species (Table OR4.1).For each of these 16 species one individual was washed in ethanol 70% and crushed in 100µl PBS.
DNA was isolated following the DNA isolation protocol of Promega (Wizard® SV Genomic DNA Purification System) with minor modifications. The 18S rRNA gene was amplified using the universal primers 4F and 3528R, a blend of Taq polymerase (MRC Holland) and pfu DNA polymerase (Promega) and using a cycling program with annealing temperature of 55°C and 30 cycles.The PCR product was purified according to the protocol of the Wizard® SV Gel and PCR Clean-Up System (Promega), ligated into the vector pGEM®-T (Promega) and cloned into XL1-blue E. coli competent cells (Stratagene).Positive clones were grown overnight in LB medium and plasmids were isolated using the Wizard® Plus SV Minipreps DNA purification System of Promega.Sequencing was done with the Big Dye v1.1 chemistry of Applied Biosystems on an ABI 3100 Avant. Besides plasmid specific primers (T7 and Sp6), we developed a set of species-specific primers to sequence the whole gene.Contigs were made per individual, using Vector NTI software, package Advance 9.0 (Invitrogen).
We used the software MUSCLE to align the sequences and Gblock procedure to exclude gaps. We used maximum likelihood (PhyML software) and Bayesian inference (MrBayes software) to generate phylogenetic trees. For maximum likelihood, we used 100 bootstrap replicates to estimate branch support. For Bayesian inference, a Markov chain was run for 10,000 generations and sampled every 10 generations to yield a posterior probability distribution of 1000 trees. The first 250 trees were discarded as “burn-in”. For both analyses, we used HKY85 nucleotide substitution model. Only branches with support values ≥ 0.5 were considered. All software was run via the website http//:
Results
Maximum likelihood and Bayesian inference resulted in almost identical phylogenetic trees, except that maximum likelihood presented a polytomy between C. convexus and the family Armadillidiidae which was resolved using Bayesian inference. Because Bayesian inference resulted in a better resolved phylogeny (Fig. OR4.1), we used this phylogenetic tree for calculations of phylogenetic signal and PIC for the functional traits. All studied traits showed a significant phylogenetic signal (surface area: VarCont =2.675, P=0.002; water loss rate: VarCont =0.006, P<0.001; desiccation resistance: VarCont=3.002, P=0.001; fatal water loss: VarCont=0.009, P=0.007). Because of the strong correspondence between phylogeny and taxonomical classification, surface area, desiccation resistance and water loss rate were also strongly associated with taxonomy. For instance, species from the family Trichoniscidae showed low values of surface area and desiccation resistance and high values of water loss rate, while species from the families Porcellionidae and Armadillidiidae showed the opposite pattern. On the other hand, fatal water loss did not show a strong link with taxonomical classification, with species from different families showing high values (e.g., Ligia oceanica, Porcellio spinicornis, Armadillidium pictum and Trichoniscoides albidus).
Fig. OR4.1. Phylogenetic relationships between 30 isopod species present in the Netherlands. The tree was constructed based on 18S rRNA. Sequences were aligned with MUSCLE and Bayesian inference of phylogeny was used to generate the tree. Only branches with support values of 0.5 or higher are shown. Species not included in the trait analyses are marked with (*). The phylogenetic tree shows a strong correspondence with the taxonomical classification.
Table OR4.1.: Accession numbers for the 18S rRNA sequences of terrestrial isopods deposited in GenBank. Accession numbers for sequences made for this study and the ones retrieved from GenBank are shown. Site of collection is also shown for species sequenced for this study.
Species / Site / Accession numberSpecies sequenced for this study
Eluma caelatum / Goes / JN232925
Metatrichoniscoides leydigii / Schoorldam / JN232926
Armadillidium album / Borselle / JN232924
Platyarthrus hoffmannseggii / Schoorldam / JN232927
Miktoniscus patiencei / Zeeland / JN232928
Haplothalmus mengii / Spierdijk / JN232929
Trichoniscoides sarsi / Spierdijk / JN232930
Trichoniscoides helveticus / Ooypolder / JN232931
Trichoniscoides albidus / Schoorldam / JN232932
Porcellium conspersum / Maastricht / JN232933
Trichoniscus provisorius / Westzaan / JN232934
Armadillidium pictum / Maastricht / JN232935
Armadillidium pulchelum / Vasse / JN232936
Trichoniscus pygmaeus / Buitenhuizen / JN232937
Trichoniscus pusillus / Atilla / JN232938
Androniscus dentiger / Maastricht / JN232939
Sequences retrieved from NCBI
Ligia oceanica / AF255698
Ligidium hypnorum / AJ287056
Haplothalmus danicus / AJ287066
Hyloniscus riparius / AJ287065
Oniscus asellus / AF255699
Philoscia muscorum / AJ287058
Cylisticus convexus / AJ287059
Trachelipus rathkei / AF279605
Porcellio spinicornis / AY048183
Porcellionides pruinosus / AY048181
Porcellio scaber / AJ287062
Armadillidium opacum / AY048176
Armadillidium nasatum / AY048175
Online Resource5: Porcellio scaber populations used as site and time controls for the trait measurements (means ± SD; untransformed data). At each site we visited to collect isopod species, P. scaber was also surveyed as a site control. Within the same week of each field trip, the P. scaber population at the garden of the Hortus Botanicus at the VU University Amsterdam was also surveyed as a time control. ANOVA tables and result of post-hoc Tukey test are shown. For the analyses with water loss rate, desiccation resistance and fatal water loss, ventral surface area, and sex were included as co-variables. Variables were log-transformed to improve normality.
Online Resource6: Alternative path models including direct effect of surface area on desiccation resistance for both (a) species averages and (b) phylogenetically independent contrasts. Although the model is supported by the data, the direct effect of surface area is not statistically significant. This indicates surface area affects desiccation resistance only through changes in water loss rate. There are no other qualitative changes in the results whencompared to the modelswe present in the article (Fig. 1).Path coefficients, goodness-of-fit index (GFI) and adjusted goodness-of-fit index (AGFI) are shown for each model (*P<0.05; **P<0.01).