Ecological Archives E093-218-A1

Zhongling Yang, Jeff R. Powell, Chunhui Zhang, Guozhen Du. 2012. The effect of environmental and phylogenetic drivers on community assembly in an alpine meadow community. Ecology 93:2321–2328. http://dx.doi.org/10.1890/11-2212.1

Appendix A. Compositional responses of plant communities, plant community responses in community-weighted mean trait value, a table of species presence or absence in alpine grassland plots associated with the studied treatments, and a figure of phylogenetic relationships among species.

Compositional responses of plant communities

A table of species present in each of the four treatments is provided in Table A1. Compositional responses of the 32 communities to fertilization and enclosure were evaluated using permutational multivariate analysis of variance (Anderson 2001) employing a modified Gower distance with a logarithmic transformation (Anderson et al. 2006). Species responses to treatments were determined using indicator species analysis (Dufrene and Legendre 1997) on the same log-transformed species table, with clusters inferred by redundancy analysis (RDA). To evaluate different clustering approaches based on the experimental treatments, we calculated the Akaike information criterion corrected for small sample size (AICc; McQuarrie and Tsai 1998) using a multivariate analogue of residual sums of squares, calculated as the sum of all eigenvalues minus the sum of the canonical eigenvalues of Y (the species table) on X (the matrix of fertilization and enclosure treatment levels), described by Legendre and Anderson (1999). These trait-independent analyses were performed in R using the ‘vegan’ (Oksanen et al. 2011) and ‘labdsv’ (Roberts 2010) packages.

Both enclosure (P = 0.001) and fertilization (P = 0.001), as well as their interaction (P = 0.001), influenced the structure of the plant communities. Visual inspection of RDA biplots (Fig. A1) suggests that communities clustered into three groups: unfertilized and ungrazed (enclosed) plots, fertilized and ungrazed plots, and grazed plots (regardless of fertilization treatment); model selection supported this interpretation (three groups: proportion constrained inertia = 0.217, AICc = 194.3; four groups: proportion constrained inertia = 0.232, AICc = 196.3). Indicator species analysis suggested that ungrazed plant communities were dominated by grasses (Poa crymophila, Elymus nutans) when fertilized, but a mix of forbs (Bupleurum malconense, Tibetia himalaica, Oxytropis kansuensis, Scirpus pumilus, Anemone obtusiloba, Potentilla fragarioides), one grass (Stipa aliena) and one sedge (Kobresia graminifolia) when unfertilized. Grazed communities featured by nine forbs (Pedicularis kansuensis, Heracleum millefolium, Leontopodium souliei, Plantago depressa, Lomatogonium carinthiacum, Potentilla anserina, Taraxacum mongolicum, Halenia elliptica, Anemone obtusiloba).

Plant community responses in community-weighted mean trait value

We looked for shifts in the mean value of trait within each treatment as a function of the environment by estimating the community-weighted mean for each trait under each environment (CWMObs) and comparing these values to those expected if trait distributions within communities assembled randomly (CWMExp) (Mason et al. 2008). In 9999 random scenarios, the species composition within communities was maintained as observed, but trait values were shuffled among all species in the same environment, all the species that had their trait values shuffled in the one average community unit representing that environment. This null model was used as it produced the most conservative results (Mason et al. 2008, Cornwell and Ackerly 2009). Therefore, CWMExpprovides a confidence interval of the abundance-weighted mean for each trait assuming that the abundances of species are unrelated to trait differences among species. P values were calculated as the proportion of CWMExp being higher/lower than CWMObs, with the p value being doubled to conform to a one-tailed test. Significance was assumed at p < 0.05. Moreover, in order to compare community-weighted mean trait values among treatments, we generated a confidence interval for CWMObs across the eight plots for each treatment. Species trait means were matched to the species present in the plots to calculate the distribution of trait values.

As expected, community-weighted mean trait values varied among the experimental treatments, but evidence of convergence within certain treatments was only observed for tissue nitrogen and plant height (Table A2). In fertilized, enclosed communities, observed values for tissue nitrogen and plant height tended to be lower and higher, respectively, than in the other treatments, and differed significantly from the expectations (Table A2), suggesting convergence within these communities with respect to these traits. Plant height was also significantly higher than expected in fertilized, grazed communities (Table A2).

TABLE A1. Species presence (1) or absence (0) in alpine grassland plots associated with the studied treatments (E0: enclosed, unfertilized; E30: enclosed, fertilized; G0: grazed, unfertilized; G30: grazed, fertilized).

 

E0

E30

G0

G30

Agrostis hugoniana

1

1

1

1

Agrostis trinii

1

1

1

1

Ajania tenuifolia

1

1

1

1

Allium sikkimense

1

1

1

1

Anaphalis lactea

1

0

1

0

Anemone obtusiloba

1

1

1

1

Anemone rivularis

1

1

1

1

Artemisia sieversiana

1

1

1

1

Astragalus polycladus

1

0

1

1

Bupleurum malconense

1

0

1

1

Carex infuscata

1

1

1

1

Delphinium kamaonense

1

1

1

1

Elymus nutans

1

1

1

1

Euphorbia altotibetica

1

1

1

1

Festuca sinensis

1

1

1

1

Galium verum

1

1

1

1

Gentianopsis barbata

1

1

1

1

Geranium wilfordii

1

1

1

1

Glaux maritima

0

0

1

0

Halenia elliptica

1

1

1

1

Heracleum millefolium

1

1

1

1

Kobresia humilis

1

1

1

1

Kobresia setchwanensis

1

1

1

1

Koeleria cristata

1

1

1

1

Lancea tibetica

1

1

1

1

Leontopodium souliei

0

0

1

1

Ligularia virgaurea

1

0

1

1

Lomatogonium carinthiacum

0

0

1

1

Medicago archiducis nicolai

1

1

1

1

Oxytropis kansuensis

1

1

1

1

Pedicularis kansuensis

0

0

1

1

Plantago depressa

1

1

1

1

Poa crymophila

1

1

1

1

Potentilla anserina

1

0

1

1

Potentilla bifurca

1

1

0

0

Potentilla fragarioides

1

1

1

1

Potentilla potaninii

1

1

1

1

Ranunculus membranaceus

1

1

1

1

Ranunculus tanguticus

1

1

1

1

Ranunculus tanguticus var capillaceus

1

1

1

1

Saussurea kansuensis

1

1

1

1

Saussurea nigrescens

1

1

1

1

Scirpus pumilus

1

1

1

1

Stellaria uda

1

1

1

1

Stipa aliena

1

1

1

1

Taraxacum mongolicum

1

1

1

1

Tibetia himalaica

1

1

1

1

Veronica eriogyne

1

1

1

1

 

TABLE A2. Observed mean community-weighted trait distributions (CWMObs, with confidence intervals in terms of null models) among plant communities exposed to the experimental treatments (E0: enclosed, unfertilized; E30: enclosed, fertilized; G0: grazed, unfertilized; G30: grazed, fertilized). Statistical tests were performed against the null expectation that species abundances within each environment were unrelated to their trait values; all tests were one-tailed. *** = P < 0.001; ** = P < 0.01; * = P < 0.05; NS = not significant.


Trait

Mean

E0

E30

G0

G30

CWMObs

95% CI

CWMObs

95% CI

CWMObs

95% CI

CWMObs

95% CI

SLA

0.85NS

0.76–1.02

0.56NS

0.49–0.78

0.88NS

0.84–1.17

0.93NS

0.87–1.12

Plant nitrogen

0.071NS

0.063–0.078

0.036**

0.038–0.053

0.075NS

0.070–0.084

0.081NS

0.078–0.094

Seed mass

0.84NS

0.41–0.93

0.53NS

0.22–0.75

0.94NS

0.46–1.03

1.01NS

0.46–0.96

Plant biomass

3.38NS

2.26–4.48

3.05NS

1.78–4.76

1.89NS

1.63–2.84

1.99NS

1.72–3.11

Plant height

106.86*

73.76–106.64

129.59**

58.28–103.41

51.91NS

38.16–57.63

65.01*

44.29–64.59

 


FigA1

 
   FIG. A1. RDA biplot of compositional shifts in alpine plant communities. Closed circles represent enclosed plots, open circles represent grazed plots; fertilized and unfertilized plots are indicated by black and gray circles, respectively. The centroids for each treatment are indicated by the position of the treatment names. Numbers in parentheses indicate the proportion of inertia accounted for by the first two constrained axes.

 

FigA2

 
   FIG. A2. Phylogenetic relationships among species in this study, generated using 'phylomatic' (Webb and Donoghue 2005) in association with the 'R20100701' version of the Angiosperm Phylogeny Group III supertree (obtained from http://www. Phylodiversity net/phylomatic/). Branch lengths from Wikstrom et al. (2001) were incorporated into the tree and undated nodes were estimated using the 'bladj' algorithm in 'phylocom' (Webb et al. 2008). Numbers correspond to the labels in Fig. 1A (in the main text).

 

LITERATURE CITED

Anderson, M. J. 2001. A new method for non-parametric multivariate analysis of variance. Austral Ecology 26:32–46.

Dufrene, M., and P. Legendre. 1997. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecological Monographs 67:345–366.

McQuarrie, A. D. R., and C. L. Tsai. 1998. Regression and time series model selection. World Scientific Publishing Company, Singapore.

Mason, N. W. H, C. Lanoiselee, D. Mouillot, J. B. Wilson, and C. Argillier. 2008. Does niche overlap control relative abundance in French lacustrine fish communities? A new method incorporating functional traits. Journal of Animal Ecology 77:661–669.

Oksanen, J., F. G. Blanchet, R. Kindt, P. Legendre, R. B. O’Hara, G. L. Simpson, P. Solymos, M. H. H. Stevens, and H. Wagner. 2011. Vegan: Community Ecology Package. R package version 1.17-7.

Roberts, D. W. 2010. LABDSV: ordination and multivariate analysis for ecology computer program, version 1.4-1.R Package.

Webb, C. O., and M. J. Donoghue. 2005. Phylomatic: tree assembly for applied phylogenetics. Molecular Ecology Notes 5:181–183.

Webb, C. O., G. S. Gilbert, and M. J. Donoghue. 2006. Phylodiversity dependent seeding mortality, size structure, and disease in a bornean rain forest. Ecology 87:S123–S131.

Webb, C. O., D. D. Ackerly, and S. W. Kembel. 2008. Phylocom: software for the analysis of phylogenetic community structure and trait evolution. Bioinformatics 18:2098–2100.

Wikstrom, N., V. Savolainen, and M. W. Chase. 2001. Evolution of the angiosperms: calibrating the family tree. Proceedings of the Royal Society B 268:2211–2220.


[Back to E093-218]