plot.habitat: Plot Options For a 'habitat' Object

In txm676/sars: Fit and Compare Species-Area Relationship Models Using Multimodel Inference

Plot Options For a 'habitat' Object

Description

S3 method for class 'habitat'. plot.habitat creates plots for objects of class habitat, using the R base plotting framework. The exact plot generated depends on whether the input data come from sar_habitat or sar_countryside.

Usage

## S3 method for class 'habitat'
plot(
  x,
  IC = "AICc",
  type = 1,
  powFit = TRUE,
  totSp = FALSE,
  xlab = NULL,
  ylab = NULL,
  lcol = NULL,
  pLeg = TRUE,
  legPos = "right",
  legInset = 0,
  ModTitle = NULL,
  which = NULL,
  ...
)

Arguments

x	An object of class 'habitat'.
IC	The information criterion weights to present (must be one of 'AIC', 'BIC' or 'AICc'), if plotting a sar_habitat object.
type	Whether a Type 1, 2, 3 or 4 plot should be generated, if plotting a sar_countryside object (see details).
powFit	For Type 1 plots, should the predicted total richness values of the power (or logarithmic) model be included as red points (logical argument).
totSp	For Type 2 and 3 plots, should a total species curve be added which is the sum of the individual species group curves for a given plot (default = FALSE).
xlab	Title for x-axis (default titles are used if not provided).
ylab	Title for y-axis (default titles are used if not provided).
lcol	For Type 2 & 3 plots: the colours of the fitted lines, for each component model. Should be a vector, the length (and order) of which should match the number of species groups in x, including the total species group (last in the order) if totSp == TRUE. If not included, randomly selected colours are used.
pLeg	For Type 2 & 3 plots: should a legend be included (logical argument), showing the line colours and corresponding species groups.
legPos	For Type 2 & 3 plots: the location of the legend. Can either be a position (e.g., "bottomright"), or the x and y co-ordinates to be used to position the legend (e.g., c(0,5)).
legInset	For Type 2 & 3 plots: the inset argument in legend. Enables the legend to be plotted outside the plotting window (it still needs the user to manually change their graphical margin parameters).
ModTitle	For Type 2 & 3 plots: a vector of plot titles, which should have the same length as the number of habitats used in the original model fit. If NULL (default), the habitat names used in the original model fit are used. For Type 4 plots: a vector of plot titles, which should have the same length as the number of species groups used in the original model fit. If NULL (default), the species group names used in the original model fit are used. If no plot titles are wanted, use ModTitle = "none".
which	For Type 2 - 4 plots: select an individual plot to generate, rather than generating the plots for all habitats. If not NULL (the default) should be a numeric vector of length 1; the order of plots matches the order of habitats / species groups in the original data used to fit the model.
...	Further graphical parameters may be supplied as arguments.

Details

The exact plot that is generated depends on the input data. If x is the fit object from sar_habitat, a simple barplot of information criterion (IC) weights for the different model fits is produced. The particular IC metric to use is chosen using the IC argument.

If x is the fit object from sar_countryside, four plot types can be produced (selected using the type argument). A Type 1 plot plots the predicted total richness values (from both countryside and Arrhenius power (or logarithmic) SAR models) against the observed total richness values, with a regression line (intercept = 0, slope = 1) included to aid interpretation.

A Type 2 plot uses countryside_extrap internally to generate separate fitted SAR curves for each of the modelled species groups, for each habitat individually, using a set of hypothetical sites (ranging in area from zero to the maximum observed site area value) in which the proportion of a given habitat is always 100 percent. See Matthews et al. (2025) for further details. A plot for each habitat is generated, unless the which argument is used to select the plot for a specific habitat. See the Examples section below.

A Type 3 plot follows a similar approach as for Type 2 plots, but instead varies the proportion of a given habitat while fixing site area. The area of the largest site in data is used, and for a given (focal) habitat, the proportion of the site represented by the focal habitat is varied from zero up to one. As site area is fixed, as the proportion of the focal habitat increases, the proportions of the other habitats decrease at an equal rate. This process is then repeated using the next habitat as the focal habitat, and so on.

Note that the logarithmic SAR model doesn't work with zero area values, so the minimum area value of the 'hypothetical' sites used to generate the fitted curves in a Type 2 or 3 plot is set to 0.01 if this model is used.

A Type 4 plot represents the effective area plots used in Merckx et al. (2019). The “effective area” for species group i in a site comprising j habitats is given by Ai=Σ hij*Aj (summed across the j habitats in the site), where hij is the affinity of species group i to habitat j and is taken from the fitted sar_countryside model. The effective area of each site in the dataset is then calculated for species group i, and the SAR plotted using the effective area values instead of standard area. If the power form of the countryside model was fitted, the effective area plot is generated in log-log space and a standard log-log power model is also generated for comparison. If the logarithmic form was used, linear-log plots (i.e, just log-transformation of area) are used instead. In both cases, a linear model is fitted to the relationships and the R2 value presented on the plots. Similarly to Type 2 and 3 plots, these plots are generated separately for each species group; the user can choose to generate the plot for a specific species group using the which argument.

References

Matthews et al. (2025) An R package for fitting multi-habitat species–area relationship models. In prep.

Merckx, T., Dantas de Miranda, M. & Pereira, H.M. (2019) Habitat amount, not patch size and isolation, drives species richness of macro-moth communities in countryside landscapes. Journal of Biogeography, 46, 956–967.

Examples

#Run the sar_habitat function and generate a barplot of the AICc
#values
data(habitat)

s <- sar_habitat(data = habitat, modType = "power_log",
con = NULL, logT = log)

plot(s, IC = "AICc", col = "darkred")

## Not run: 
#Run the sar_countryside function and generate a Type 1 plot,
#including the predicted values of the standard power model
data(countryside)

s3 <- sar_countryside(data = countryside, modType = "power",
gridStart = "partial", habNam = c("AG", "SH",
"F"), spNam = c("AG_Sp", "SH_Sp", "F_Sp", "UB_Sp"))

plot(s3, type = 1, powFit = TRUE)

#Generate Type 2 plots providing set line colours, plot titles,
#and modifying other aspects of the plot using the standard
#base R plotting commands.

 plot(s3, type = 2, lcol = c("black", "aquamarine4",
"#CC661AB3" , "darkblue"), pLeg = TRUE, lwd = 1.5, 
 ModTitle = c("Agricultural land", "Shrubland", "Forest"))
 
#Generate the same plots, but all in a single plotting window,
#using the ask argument
 par(mfrow = c(2, 2))
 plot(s3, type = 2, lcol = c("black", "aquamarine4",
"#CC661AB3" , "darkblue"), pLeg = FALSE, lwd = 1.5, 
 ModTitle = c("Agricultural land", "Shrubland", "Forest"),
 ask = FALSE)
 
 dev.off()
 
#Select a single plot to generate, including
#a legend and positioning it outside the main plotting window.
#Note this will change the graphical margins of your plotting
#window.
par(mar=c(5.1, 4.1, 4.1, 7.5), xpd=TRUE)

plot(s3, type = 2, lcol = c("black", "aquamarine4",
"#CC661AB3" , "darkblue"), pLeg = TRUE,  legPos ="topright",
legInset = c(-0.2,0.3), lwd = 1.5, ModTitle = "Forest",
which = 3)

dev.off()

#Generate Type 3 plots (here only displaying the first), including
#a total species richness curve
plot(s3, totSp = TRUE, type = 3, lcol = c("black",
"aquamarine4", "#CC661AB3" , "darkblue", "darkgrey"), pLeg =
TRUE, lwd = 1.5, ModTitle = c("Agricultural land",
"Shrubland", "Forest"), which =1)

#Generate Type 4 plots (here only displaying the first)
plot(s3, type = 4, which =1)


## End(Not run)

txm676/sars documentation built on April 13, 2025, 3:59 p.m.