Variations in peat bog spider communities related to environmental heterogeneity

David J. Curtis & Eric Bignal

Back to Arachnologia  Return to Arachnologia

Adapted, August 2002, for presentation as a web page, 
from the original paper which should be cited as:
Curtis, D.J. & Bignal, E.M. (1980) Variations in peat-bog spider communities in relation to environmental heterogeneity.
Proc. 8th. Int. Congr. Arachnol., Vienna, 81-86.


 

The composition and structure of araneid/opilionid communities depends upon environmental conditions in complex ways. Firstly, habitat properties determine the distribution and thus the probability of co-occurrence of species to form a community. Secondly, community structure is the expression of component species' interactions, mostly competitive, which are influenced by environmental patchiness and the distribution of species over microhabitats. In this paper we are concerned with data illustrating the influence of environmental heterogeneity on the ground-layer spider (and harvest-spider) communities of peat bogs.

Vegetation physiognomy is a complex indicator of habitat conditions. The physical structure of vegetation is important in itself, but it also depends upon and influences microhabitat conditions. Plant tussocks or the dwarf shrubs which are found on peat bogs provide shelter with relatively constant microclimates affected by the height, density and orientation of the vegetation. Vegetation physiognomy on the peat bogs is also related to many other aspects, such as general climatic conditions, degree of exposure, peat drainage, and other effects of management, e.g. periodic moor-burning. Phytosociological characteristics could also be taken into account and will be considered elsewhere.

Sample sites

Eleven peatland sites in the Strathclyde Region of western Scotland were sampled over one year from spring 1977 to spring 1978. The locations of these sites and their vegetation are listed elsewhere (1). They are generally dominated by dwarf shrubs: ubiquitously Calluna vulgaris, often Erica tetralix, occasionally Vaccinium myrtillus, together with grasses such as Deschampia flexuosa and cotton-grasses Eriophorum angustifolium and E. vaginatum.

Single sample stations were placed on Blood Moss, Dalmellington Moss and Longriggend Moss, but at the partially wooded bogs samples were taken in both open and wooded areas. Carnwath Moss and Wood, Braehead Moss and Wood, North Bellstane Plantation (NBP) Moor and Wood, Bloak Moss and Wood. Birch woodland is present at Braehead and NBP Woods, the latter having a very wet floor and relatively lush vegetation. In contrast, Bloak Wood is an anthropogenic Scots pine plantation with ground flora formed by degeneration of the peat-bog community drying out. At Carnwath the Wood site is in sparse birch and pine on a drier area of the bog with degeneration and consequent patchiness of the heather.

Ten pitfall traps with formalin/detergent trapping fluid were set in a 1 m2 quadrat at each sampling station. These were emptied at varying intervals and data combined over the whole year; temporal variation is not considered here. Environmental variability of each site was assessed by examining the vegetation at 1 m intervals over a 5 x 5 m grid centred on the pitfall station.

Environmental description

Vegetation physiognomy was surveyed using a vertical quadrat method (2) providing 25 data points at each site. Variables recorded were the frequency of vertical, and of horizontal vegetation features, together with corresponding density, within 10cm intervals from the ground up to 1 m. The combined data for all 275 sample points were subjected to a principal components analysis (PCA; of the covariance matrix) with % cover density values scaled by 0.25 to accord with the frequency range of 0-25. This analysis provided three components accounting for nearly 75% of total variation and representing vegetation abundance, 'shrubbiness' and orientation.
 
 
 
 
 
Table 1: Vegetation physiognomy; relative contribution of features.

Contributors

Principal Components

Features

Height Range (cm)

I

II

III
Vertical 0 - 10 0 -0.457 0.558
10 - 20 0.375 -0.316 0.196
20 - 30 0.449 0.179 0
30 - 40 0.263 0.356 0.184
40 - 50 0.113 0.256 0.235
50 - 60 0 0.122 0
Horizontal 0 - 10 0.188 0 -0.551
10 - 20 0.246 0 -0.406
20 - 30 0.184 0.103 0
30 - 40 0 0.136 0
Density 0 - 10 0.238 -0.443 0
10 - 20 0.409 -0.234 -0.122
20 - 30 0.383 0.152 0
30 - 40 0.209 0.258 0
40 - 50  0 0.207 0.161
% of total variance:- 43.8 21.5 9.4
Interpretation:- abundance shrubbiness orientation
all non-zero values
are positive		 contrast negative
or zero values near
ground with positive
values higher up 		 contrast positive
values for vertical
features with negative
values for horizontal

 

The PCA plot in Fig. 1 indicates the relative positions of the sites, in terms of vegetation structure, and their different amounts of variation. The latter may be quantified by the trace (T) of the covariance matrix, i.e. the summed variances of all variables. The volume (V) enclosed by an ellipsoid described by taking one standard deviation to either side of a site's centroid of PCA scores may be used to express the relative spread of the site in a hyperspace defined by the PCA axes. Values of T and V are given in Table 2; they represent measures of environmental heterogeneity.
 

PCA plot - vegetation
 
 
 
 

Fig. 1. Vegetation physiognomy PCA.

Above: plot of axes I and II.
Letters indicate the centroids of scores for 25 samples at each site A - K and are surrounded by lines drawn to apices at one standard deviation either side of mean on each axis.
Arrows indicate increasing vegetation density a1ong axis I and increasing shrubbiness along axis II,

Below: sites along axis III.
Arranged from predominantly horizontal to mainly vertical orientation of structures. Lengths of lines drawn proportional to standard deviation of scores at each site.
 
 
 
 
 
 
 
 
 

Table 2: Values of environmental and community parameters for the 11 sites,
arranged in ascending order of environmental heterogeneity.
Site T V N S E(S) B
A Braehead Wood 199 915 939 59 28.8 16.48
B Longriggend Moss 233 1401 1247 60 26.0 10.12
C Dalmellington Moss 323 2041 686 45 23.8 7.80
D Blood Moss  339 1833 886 35 16.1 6.46
E Carnwath Moss 356 2046 1036 61 28.6 15.17
F Braehead Moss  516 4355 862 54 29.6 11.95
G NBP Wood  641 3972 989 72 37.0 20.95
H Bloak Wood  699 6721 790 65 31.7 13.58
I Bloak Moss 733 5764 401 49 30.3 12.37
J NBP Moor 949 10189 1147 63 30.2 14.92
K Carnwath Wood 996 10085 867 66 35.4 26.00

Community description

Table 2 also gives values for community parameters for the sites, derived from the total sample of 9850 individuals comprising 161 species. N, the number of individuals captured, includes immatures which are excluded from the other parameters. Species richness, S, is the number of species recorded. As N varies greatly between sites, the number of species, E(S), expected in a standard sample size of 100 adults was calculated (3). Species diversity, B, was calculated as the reciprocal of dominance, i.e. B = 1 / Σ pi2 , where pi is the proportional representation of species i at the site.

A PCA plot of the spider data is shown in Fig. 2, on which deduced gradients are indicated. PCA plot - spiders
The first three axes accounted for 32.6%, 20.3% and 11.9% respectively of the total variance.
 
 

Fig. 2: Spiders PCA.
Plot of axes I and II.
From the group of mostly wooded sites (encircled)
trend 1 is to wet, exposed sites
and trend 2 to open, dry sites.

 
 




Discussion is here confined to the more abundant species and the distribution of individual species will be considered fully elsewhere. Relative contributions made by the various species indicates groupings:- (a) a woodland or litter-dwelling group includes opilionids plus Diplocephalus picinus (Blackwall 1841), Lepthyphantes zimmermanni Bertkau 1890, Pardosa nigriceps (Thorell 1856), Pirata uliginosus (Thorell 1856), Gonatium rubens (Blackwall 1833) and Lepthyphantes tenebricola (Wider 1834); (b) Alopecosa pulverulenta (Clerck 1757) and Pachygnatha degeeri are associated with dry, open sites, while (c) Pardosa pullata (Clerck 1757), Pirata piraticus (Clerck 1757) and Centromerita concinna Thorell 1875 tend to wetter, open sites; (d) Lepthyphantes mengei Kulczynski 1887, Cnephalocotes obscurus (Blackwall 1834), Trochosa terricola Thorell 1856 and Antistea elegans are intermediate between group (a) and groups (b) and (c).
 

Species frequency curves
 
 

The value of E(S) depends upon the frequency of species represented by different numbers of individuals and species frequency curves are drawn in Fig. 3. Variations are apparent from one site to another.

  Fig. 3. Species frequency curves
for sites A - K with E(S) values indicated.
Abscissa scale is in octaves: l, 2, 4, 8, 16, 32, 64, 128, 256.
All to the same scale.
Note the trend for increasing frequency (f) of rare species.

 








Correlations between community structure and environment.

The trend for higher frequencies of rare species at sites of greater environmental variation is reflected in the corresponding rise in E(S) values. Clear correlation is apparent between E(S) and V or T (Table 3).
 
 
 
Table 3: Values of Spearman rank correlation coefficient (above diagonal) and corresponding significant probabilities (below).
T V N S E(S) B
T   0.964 -0.218 0.509 0.746 0.427
V 0.000   -0.277 0.491 0.691 0.355
N - -   0.382 -0.136 0.236
S - - -   0.782 0.809
E(S) 0.008 0.018 - 0.005   0.746
B - - - 0.003 0.008  

There is good agreement between the community parameters (highlighted yellow)  and also between V and T (highlighted green); N does not correlate with anything. Without the two birch-wood sites, there is a correlation of 0.683 between S and V (p = 0.041), i.e. between species richness and environmental heterogeneity. The pink cells in Table 3 highlight the significant correlation between the "expected number of species" and environmental heterogeneity.

Relatively few species were recorded at site D, at a higher altitude and more exposed than other sites, but relatively many at the two birch-wood sites, A and G, influenced by the occasional migration of canopy and field-layer species to the ground.

Because of the mobility of spiders, it is difficult to distinguish 'between-habitat' from 'within-habitat' diversity. In other words, are we measuring point-, or α-diversity, sensu Whittaker (9), or is it so affected by movement of species from adjacent (micro-) habitats that it represents β-diversity? A relationship between species diversity and spatial variation has been hypothesised (7) for spider communities and demonstrated for many other groups of animals. Increased environmental heterogeneity permits the coexistence of more species (6,8) and may influence the competitive interactions in the community (5).

Pianka (4, p.292) neatly lists ten mechanisms which may influence species diversity. Given that these interact with each other, it is clear that the data reported here agree with the spatial heterogeneity hypothesis, but probably are not entirely explained by it alone.

References

(1) CURTIS, D.J. (1985) Scottish wetland spiders. I. Peat-bogs in Strathclyde. Scottish Naturalist, 1985: 45-65.

(2) CURTIS, D.J. (1980): A simple method for obtaining quantitative information on vegetation physiognomy. Scottish Field Studies, 1980, 5 - 11.

(3) HECK Jr., K.L , VAN BELLE, C. & SIMBERLOFF, D. (1975): Explicit calculation of the rarefaction diversity measurement and the determination of sufficient sample size. Ecology 56, 1459-1461.

(4) PIANKA, E.R. (1978): Evolutionary Ecology. New York (Harper & Row).

(5) POST, W.M. & RIECHERT, S.E. (1977): Initial investigations into the structure of spider communities. I. Competitive effects. J. Anim. Ecol. 46, 729-749.

(6) TURNER, M. & POLlS, C.A. (1979): Patterns of coexistence in a guild of raptorial spiders. J. Anim. Ecol. 48, 509-520.

(7) UETZ, G.W. (1975): Temporal and spatial variation in species diversity of wandering spiders (Araneae) in deciduous forest litter. Environ, Entomol. 4, 719-724.

(8) UETZ, G.W. (1977): Coexistence in a guild of wandering spiders. J. Anim, Ecol. 46, 531-541.

(9) WHITTAKER, R.H. (1969): Evolution of diversity in plant communities. Brookhaven Symp. Biol. 27, 178-196.

Addresses (updated):

Prof. D.J. Curtis, Biological Sciences, University of Paisley, High Street, Paisley, Scotland.

Dr. E. Bignal, c/o Joint Nature Conservation Committee, Northminster House, Peterborough, UK.

.

Return to..? Back to Arachnologia  Arachnologia
Ariadne's thread  Ariadne Home Page