Learn R Programming

bipartite (version 2.16)

plotweb: Visualize a bipartite interaction matrix (e.g. a foodweb)

Description

A two dimensional matrix is plotted as a bipartite graph.

Usage

plotweb(web,
    method = "cca", empty = TRUE, labsize = 1, ybig = 1,  y.width.low = 0.1, 
    y.width.high = 0.1, low.spacing = NULL, high.spacing = NULL,
    arrow="no",  col.interaction="grey80", col.high = "grey10", 
    col.low="grey10",  bor.col.interaction ="black", bor.col.high="black", 
    bor.col.low="black", high.lablength = NULL, low.lablength = NULL,
    sequence=NULL, low.abun=NULL, low.abun.col="green", 
    bor.low.abun.col ="black", high.abun=NULL, high.abun.col="red", 
    bor.high.abun.col="black", text.rot=0, text.high.col="black", 
    text.low.col="black", adj.high=NULL, adj.low=NULL, plot.axes = FALSE,
    low.y=0.5, high.y=1.5, add=FALSE, y.lim=NULL, x.lim=NULL, low.plot=TRUE, 
    high.plot=TRUE, high.xoff = 0, low.xoff = 0, high.lab.dis = NULL, 
    low.lab.dis = NULL, abuns.type="additional")

Arguments

web

Web is a matrix representing the interactions observed between higher trophic level species (columns) and lower trophic level species (rows). Usually this will be number of pollinators on each species of plants or number of parasitoids on each species of prey.

method

Default method is cca, which leads to as few crossings of interactions as possible. The other option is normal, which leaves order as given by the matrix.

empty

logical; should empty columns or empty rows be omitted from plotting; defaults to true

labsize

factor for size of labels, default is 1

ybig

vertical distance between upper and lower boxes, default is 1

y.width.low

width of lower boxes, default is 0.1

y.width.high

width of upper boxes, default is 0.1

low.spacing

distance between lower boxes, default is NULL, so automatically spaced that length of upper and lower boxes is the same. Be aware if set to any value that x.lim may has to be adjusted to ensure that the network is not plotted outside the plotting region

high.spacing

distance between upper boxes, default is is NULL, so automatically spaced that length of upper and lower boxes is the same. Be aware if set to any value that x.lim may has to be adjusted to ensure that the network is not plotted outside the plotting region

arrow

display type of connection between upper and lower boxes, options are up, down, both, up.center, down.center, both.center and no, default is no, which is a polygonal connection between boxes.

col.interaction

color of interaction, default is grey80.

col.high

color of upper boxes, default is grey10.

col.low

color of lower boxes, default is grey10.

bor.col.interaction

border color of interaction, default is black

bor.col.high

border color of upper boxes, default is black

bor.col.low

border color of lower boxes, default is black

high.lablength

number of characters of upper labels that should be plotted. If zero no labels are shown, default is NULL which plots the complete labels.

low.lablength

number of characters of lower labels that should be plotted. If zero no labels are shown, default is NULL which plots the complete labels.

sequence

list of two with two names vectors: seq.high and seq.low, which specify the order in which species are plotted. Cannot be set for method="cca". Defaults to NULL, where the sequence remains as given or is determined by the CCA internally.

low.abun

Named vector with independent abundance estimates for the lower trophic level, NULL if none exists. See Notes!

low.abun.col

Colour for depicting the abundance estimates for the lower trophic level; defaults to green.

bor.low.abun.col

border color for depicting the abundance estimates for the lower trophic level, default is black

high.abun

Named vector with independent abundance estimates for the higher trophic level, NULL if none exists. See Notes!

high.abun.col

Colour for depicting the abundance estimates for the lower trophic level; defaults to red.

bor.high.abun.col

border color for depicting the abundance estimates for the higher trophic level, default is black

text.rot

orientation of labels in the plot (to avoid overlapping of horizontal labels if dimension of the webs are high), default is 0 for horizontal labels, use text.rot=90 for vertical labels.

text.high.col

Colour for text labels of higher trophic level, a vector of colours can be given

text.low.col

Colour for text labels of lower trophic level. A vector of colours can be given

adj.high

Adjust upper labels. See adj in ?text how to adjust labels

adj.low

Adjust upper labels. See adj in ?text how to adjust labels

plot.axes

axis are plotted. Sometimes useful to place boxes in higher order plots. Defaults to FALSE

low.y

Position of the higher boxes on the y-axis. Defaults to 1.5

high.y

Position of the higher boxes on the y-axis. Defaults to 1.5

add

if set to TRUE a new bipartite network is added to the previous plot. In this way multitrophic webs can be plotted, see examples below. Defaults to FALSE.

y.lim

To set limits for y-axis. Useful if labels are plotted outside the plotting region and for multitrophic plots, see examples below

x.lim

To set limits for x-axis. Useful if labels are plotted outside the plotting region and for multitrophic plots, see examples below

low.plot

Defines if lower boxes should be drawn. Use in multitrophic plots to avoid plotting boxes of some trophic levels - see examples below. Defaults to TRUE

high.plot

Defines if higher boxes should be drawn. Use in multitrophic plots to avoid plotting boxes of some trophic levels - see examples below. Defaults to TRUE

high.xoff

allows to set an offset to upper boxes. Useful if high.spacing is used to centre boxes manually. Use plot.axes=TRUE for easy centring

low.xoff

allows to set an offset to lower boxes. Useful if low.spacing is used to centre boxes manually. Use plot.axes=TRUE for easy centring

high.lab.dis

normally labels are staggered to avoid plotting over themselves. if set to 0, higher labels are all on one horizontal line. By using ad.low the position of the labels can be adjusted. If set to any other value labels are staggered with this distance. Defaults to NULL

low.lab.dis

normally labels are staggered to avoid plotting over themselves. if set to 0, lower labels are all on one horizontal line. By using ad.low the position of the labels can be adjusted. If set to any other value labels are staggered with this distance. Defaults to NULL

abuns.type

How to plot abundances - are they independent (e.g. flower cover) measurements or are they additional (e.g. unparasitised hosts) measurements? Defaults to additional. Option none is interpreted in the same way as additional. See Notes!

Value

Returns a window with a bipartite graph of a food web. For all colours vectors can be used (which are recycled if length differs. Now more trophic webs can be plotted by using plotweb and the add switch, which allows to add more webs and staggering them on top of each other. Preferred option is here to order webs by yourself and use method="normal" to keep your preferred order. See examples on three and four trophic networks.

References

Tylianakis, J. M., Tscharntke, T. and Lewis, O. T. (2007) Habitat modification alters the structure of tropical host-parasitoid food webs. Nature 445, 202--205

See Also

For a different plot of food webs see visweb

Examples

Run this code
# NOT RUN {
data(Safariland)
plotweb(Safariland)

# shorter labels
plotweb(Safariland, high.lablength=3, low.lablength=0, arrow="down")

# centered triangles for displaying interacions
plotweb(Safariland, text.rot=90, arrow="down.center", col.interaction="wheat2",
	y.lim=c(-1,2.5))

#orginal sequence, up arrows and different box width
plotweb(Safariland, method="normal", arrow="up", y.width.low=0.3, low.lablength=4)
# interactions as lines
plotweb(Safariland, arrow="both", y.width.low=0.05, text.rot=90, col.high="blue", 
	col.low="green")

# add an abundance vector for lower trophic species 
low.abun = round(runif(dim(Safariland)[1],1,40)) #create
names(low.abun) <- rownames(Safariland)
plotweb(Safariland, text.rot=90, low.abun=low.abun, col.interaction="purple", 
	y.width.low=0.05, y.width.high=0.05)

plotweb(Safariland, text.rot=90, low.abun=low.abun, col.interaction ="red", 
	bor.col.interaction="red", arrow="down")

# now vectors for all colours can be given, to mark certain species or 
# interactions. Colour vectors are recycled if not of appropriate length
plotweb(Safariland,col.high=c("orange","green"))
plotweb(Safariland,col.low=c("orange","green"),col.high=c("white","grey","purple"),
	text.high.col=c("blue","red"), col.interaction=c("red",rep("green",26),rep("brown",242)),
	bor.col.interaction=c(rep("green",26),rep("brown",242)),method="normal", 
	text.rot=90, low.lablength=10, high.lablength=5)


#example one (tritrophic)
plotweb(Safariland,y.width.low=0.1, y.width.high=0.05,method="normal", 
	y.lim=c(0,3), arrow="up", adj.high=c(0.5,1.5), col.high="orange",
	high.lablength=3,high.lab.dis=0)

plotweb(t(Safariland), y.width.low=0.05, y.width.high=0.1, method="normal",
	add=TRUE,low.y=1.5,high.y=2.5, col.low="green", text.low.col="red", 
	low.lab.dis=0, arrow="down", adj.low=c(0.5,1.1),low.plot=FALSE)

#example two (4 trophic with abundance)
low.abun = round(runif(dim(Safariland)[1],1,40)) #create
names(low.abun) <- rownames(Safariland)
plotweb(Safariland, text.rot=90, high.abun=low.abun, col.interaction="purple", 
	y.lim=c(0,4.5), high.lablength=0, arrow="up", method="normal", 
	y.width.high=0.05)

plotweb(t(Safariland), y.width.low=0.05, y.width.high=0.1, method="normal", 
	add=TRUE, low.y=1.7,high.y=2.7, col.low="green", text.low.col="black", 
	low.lab.dis=0, arrow="down", adj.low=c(0.5,1.1), low.lablength=4, 
	high.lablength=0)

plotweb(Safariland,y.width.low=0.05, y.width.high=0.1, method="normal", 
	add=TRUE, low.y=2.95, high.y=3.95, col.low="green", text.low.col="black", 
	low.lab.dis=0, arrow="down", adj.low=c(0.5,1.1), low.lablength=4)

# now some examples with the abuns.type-option:
plotweb(Safariland, abuns.type='independent',arrow="down.center")
plotweb(Safariland, abuns.type='additional',arrow="down.center")

# }

Run the code above in your browser using DataLab