Appendix A. Niche breadth calculations and properties.
Although several other measures are available to quantify habitat specialization (i.e., niche breadth in our context) such as species distribution along environmental gradients (e.g., Dolédec et al. 2000) or species co-occurrences data (Fridley et al. 2007), we explicitly chose the Levins’ approach because it determines habitat specialization based on a particular measure of niche breadth (B). Here B is a function of uniformity of distribution of species abundance among the resource states for a community at hand (Levins 1968, Colwell and Futuyma 1971). This method assigns different niche breadths even for species that have the same overall abundance in the landscape of patches if one species is uniformly distributed (generalist) while the other shows clumped distribution (specialist). Further, B is independent of the environmental, biological and spatial variables measured for the rock pool metacommunity because we defined “resource state” as one habitat patch or rock pool in this study. The calculated niche breadth is not directly related to the environmental variables within the pools, or the spatial location of the different pools relative to each other in the landscape. This is a necessary condition to avoid circular arguments.
Thus this method is most suitable given our aim to assess the effects of spatial and environmental variables on metacommunity dynamics. Other methods are burdened with features that might compromise our inferences. The method we chose is commonly used in many ecological studies to quantify habitat specialization (e.g., Feldhamer et al. 1993, Fugi et al. 2008). Finally, to verify that the Levins’ measure adequately captures species specialization, we calculated niche breadth based on environmental variables alone using Canonical Correspondence (CCA) and found it was reasonably well correlated with the former. While we cannot use the CCA based niche breadth in our variance partitioning analysis for the reasons stated above, the convergence of the two methods reinforces the suitability of the Levins’ measure.
FIG. A1. The relationship between tolerances (root mean square standard deviation across four axes, RMSTOL) and niche breadth calculated by Levins’ method (1968). The RMSTOL was obtained for each species using CANOCO’s Canonical Correspondence Analysis from four significant CCA axes. |
TABLE A1. Descriptive statistics of the data set. AD represents average density of each species across all years and pools, SE: standard error; Min: minimum density, Max: maximum density and NB: Niche breadth.
Species name |
AD |
SE |
Min |
Max |
NB |
Dipteran sp. |
0.15 |
0.06 |
0 |
14 |
6.44 |
Polychaete T |
0.62 |
0.22 |
0 |
46 |
7.61 |
Leidigia leidigi |
1.91 |
0.59 |
0 |
91 |
10.13 |
Dorsovilleid polychaete |
1.99 |
0.43 |
0 |
111 |
5.3 |
Metis sp. |
2.1 |
0.77 |
0 |
164 |
7.32 |
Paracyclops fimbriatus (Fischer) |
2.49 |
0.48 |
0 |
73 |
24.42 |
Tanypodid sp. |
3.06 |
0.86 |
0 |
191 |
12.08 |
Culex sp. |
3.38 |
0.56 |
0 |
71 |
32.52 |
Gyratrix hermaphroditus |
3.76 |
1.2 |
0 |
322 |
9.57 |
Copepod sp. 2 |
4.45 |
2.81 |
0 |
854 |
2.5 |
Candona sp. |
6.02 |
1.48 |
0 |
227 |
15.84 |
Sesarma miersi Rathburn |
6.23 |
1.03 |
0 |
149 |
32.72 |
Oligochaete sp. |
9.3 |
2.89 |
0 |
636 |
10.08 |
Cypricercus sp. |
11.11 |
3.45 |
0 |
803 |
10.1 |
Cytheromorpha sp.1 |
12.65 |
3.86 |
0 |
831 |
10.39 |
Heterocypris sp. |
19.16 |
5.78 |
0 |
1325 |
10.66 |
Copepod sp. |
30.5 |
14.9 |
0 |
3986 |
4.14 |
Alona davidii |
34.7 |
10.2 |
0 |
2160 |
11.31 |
Nematode sp. |
44.9 |
27.1 |
0 |
8390 |
2.72 |
Ceriodaphnia rigaudi Richard |
61 |
12.8 |
0 |
1895 |
21.21 |
Cypridopsis cf. mariae Rome |
62.2 |
13.7 |
0 |
2004 |
19.26 |
Potamocypris sp. |
84.2 |
20.9 |
0 |
3243 |
15.48 |
Orthocyclops modestus (Herrick) |
92.9 |
23.4 |
0 |
6195 |
15.12 |
Nitocra spinipes Boeck |
293.1 |
59.2 |
0 |
10845 |
22.81 |
TABLE A2. Variance partitioning among environmental and spatial variables for the three sets of data: all 24 common species (all species), habitat specialists, and habitat generalists for each of the nine years of study of the 49 rock pools located at the DBML. Abbreviations: E = variation explained by environmental variables; S = variation explained by spatial variables; E|S = variation explained by pure environmental variables (contribution of S removed); and S|E = variation explained by pure spatial variables (contribution of E removed). UV = unexplained variation, EøS = variation shared by environmental and spatial variables. The significance was determined from Monte Carlo permutations test under the reduced model from partial RDA.
|
|
Variance explained (%) |
|||||||||
Years |
Groups |
E |
p |
S |
p |
E|S |
p |
S|E |
p |
EøS |
UV |
1989 Dec. |
All species |
31.1 |
0.00 |
41.9 |
0.00 |
14.7 |
0.47 |
25.5 |
0.00 |
16.4 |
43.4 |
1990 Jan. |
30.1 |
0.00 |
45.1 |
0.00 |
14.1 |
0.56 |
29.1 |
0.00 |
16.0 |
40.8 |
|
1991 Jan |
28.2 |
0.00 |
37.7 |
0.00 |
15.7 |
0.02 |
25.2 |
0.00 |
12.5 |
46.6 |
|
1992 Jan. |
30.1 |
0.00 |
45.1 |
0.00 |
14.1 |
0.19 |
29.2 |
0.00 |
16.0 |
40.7 |
|
1993 Jan. |
30.1 |
0.00 |
32.0 |
0.00 |
16.6 |
0.02 |
18.5 |
0.07 |
13.5 |
51.4 |
|
1994 Jan. |
19.1 |
0.11 |
30.8 |
0.00 |
13.6 |
0.48 |
25.3 |
0.02 |
5.5 |
55.6 |
|
1997 Jan. |
38.2 |
0.00 |
38.5 |
0.00 |
17.4 |
0.03 |
17.8 |
0.02 |
20.8 |
44.0 |
|
1997 Jun. |
31.1 |
0.00 |
46.9 |
0.00 |
10.5 |
0.82 |
26.2 |
0.02 |
20.6 |
42.7 |
|
1998 Jan. |
32.5 |
0.00 |
43.4 |
0.00 |
10.8 |
0.39 |
21.8 |
0.00 |
21.7 |
45.7 |
|
Average |
30.1 |
|
40.2 |
|
14.2 |
|
24.3 |
|
15.9 |
45.7 |
|
1989 Dec. |
Habitat generalists |
30.7 |
0.00 |
47.0 |
0.00 |
12.5 |
0.48 |
28.3 |
0.00 |
18.2 |
41.0 |
1990 Jan. |
27.3 |
0.00 |
46.6 |
0.00 |
7.9 |
0.76 |
27.1 |
0.00 |
19.4 |
45.6 |
|
1991 Jan |
32.1 |
0.00 |
44.5 |
0.00 |
10.7 |
0.20 |
20.1 |
0.05 |
21.4 |
47.8 |
|
1992 Jan. |
32.4 |
0.01 |
55.2 |
0.00 |
14.3 |
0.19 |
37.1 |
0.00 |
18.1 |
30.5 |
|
1993 Jan. |
31.5 |
0.00 |
30.9 |
0.00 |
20.0 |
0.02 |
29.3 |
0.01 |
11.5 |
39.2 |
|
1994 Jan. |
18.7 |
0.11 |
35.4 |
0.00 |
14.7 |
0.23 |
33.9 |
0.00 |
4.0 |
47.4 |
|
1997 Jan. |
42.9 |
0.00 |
49.9 |
0.00 |
10.2 |
0.38 |
17.2 |
0.17 |
32.7 |
39.9 |
|
1997 Jun. |
29.1 |
0.01 |
50.1 |
0.00 |
8.6 |
0.86 |
29.5 |
0.03 |
20.5 |
41.4 |
|
1998 Jan. |
36.3 |
0.00 |
44.5 |
0.00 |
11.4 |
0.32 |
19.6 |
0.07 |
24.9 |
44.1 |
|
Average |
31.2 |
|
44.9 |
|
12.3 |
|
26.9 |
|
19.0 |
41.9 |
|
1989 Dec. |
Habitat specialists |
44.5 |
0.00 |
23.1 |
0.17 |
35.2 |
0.00 |
14.4 |
0.24 |
9.3 |
41.1 |
1990 Jan. |
45.1 |
0.00 |
23.6 |
0.15 |
35.0 |
0.00 |
13.4 |
0.30 |
10.1 |
41.5 |
|
1991 Jan |
26.3 |
0.01 |
24.1 |
0.19 |
22.0 |
0.02 |
19.8 |
0.24 |
4.3 |
53.9 |
|
1992 Jan. |
40.1 |
0.00 |
22.1 |
0.70 |
31.7 |
0.02 |
13.7 |
0.89 |
8.4 |
46.2 |
|
1993 Jan. |
37.8 |
0.00 |
21.6 |
0.41 |
36.9 |
0.00 |
19.7 |
0.09 |
0.9 |
42.5 |
|
1994 Jan. |
30.5 |
0.01 |
32.7 |
0.00 |
12.8 |
0.56 |
10.6 |
0.95 |
17.7 |
58.9 |
|
1997 Jan. |
65.9 |
0.00 |
29.3 |
0.05 |
44.9 |
0.00 |
24.1 |
0.05 |
21.0 |
10.0 |
|
1997 Jun. |
71.9 |
0.00 |
30.9 |
0.04 |
48.4 |
0.00 |
27.5 |
0.04 |
23.5 |
0.6 |
|
1998 Jan. |
64 |
0.00 |
23.7 |
0.16 |
48.6 |
0.00 |
8.3 |
0.42 |
15.4 |
27.7 |
|
Average |
47.3 |
|
25.7 |
|
35.1 |
|
16.8 |
|
12.3 |
35.8 |
LITERATURE CITED
Colwell, R. K., and D. J. Futuyama. 1971. On the measurement of niche breadth and overlap. Ecology 52:567–576.
Dolédec S., D. Chessel, and C. Gimaret-Carpentier. 2000. Niche separation in community analysis: a new method. Ecology 81:2914–2927.
Feldhamer, G. A., R. S. Klann, A. S. Gerard, and A. C. Driskell. 1993. Habitat partitioning, body size, and timing of parturition in pygmy shrews and associated soricids. Journal of Mammalogy 74:403–411.
Fridley, J. D., D. Jason, D. Vandermast, D. Kuppinger, M. Dane, M. Manthey, R. K. Robert. 2007. Co-occurrence based assessment of habitat generalists and specialists: a new approach for the measurement of niche width. Journal of Ecology 95:707–722.
Fugi. R, K. D. G. Luz-Agostinho, and A. A. Agostinho. 2008. Trophic interaction between an introduced (peacock bass) and a native (dogfish) piscivorous fish in a Neotropical impounded river. Hydrobiologia 607:143–150.
Levins, R. 1968. Evolution in Changing Environments. Princeton University Press, Princeton, New Jersey, USA.