Tropical Forest Community Ecology

Explaining Geographic Range Size by Species Age 49

Range size

Range size

Species age

(a)

(b)

Species age

Figure 4.1 Graphical depictions of range-size
trajectories of species over time. Black dots indicate
sampling points in time. Ideally, fossil analyses can
allow the range size of a species to be assessed at
multiple time points (a), effectively sampling over the
lifespan of a species. When using molecular estimates
of ages, species can usually be sampled at only a single
point in time (b). By sampling multiple species (different
lines on the graph), a general relationship between
species age and range size can be inferred. However,
even if all species show roughly the same shape for an
age and area relationship (e.g., hump-shaped), if they
follow varying range-transformation trajectories,
sampling single points over time will introduce
considerable variation into the species age and range
size relationshi pand make inferring general trends
more difficult.

obscured when clades with distinct evolutionary

histories are combined.

AN EMPIRICAL TEST USING A

TROPICAL PLANT GENUS

Willis developed the age-and-area hypothesis

thinking about tropical floras, and even his critics

acknowledged that the hypothesis might be more

important in the tropics (Gleason 1924), which

were seen as stable and relatively homogeneous.

Despite this early attention to the tropics, to our

knowledge there have been no explicit tests of

the hypothesis using tropical plants.The immense

diversity of tropical plant species is only beginning

to receive a genetic treatment, and our estimates

of species’ range sizes are imperfect, but slowly

improving (e.g., Pitmanet al. 2001). Most of

the molecular dating of tropical plants to date

has been conducted at higher phylogenetic lev-

els; typically these studies are concerned with the

general age of families and genera, and inferring

when and where these groups of species diversi-

fied (e.g., Daviset al. 2005, Zeregaet al. 2005,

Muellneret al. 2006). In contrast, analyses of

age and area require species-level resolution to

properly address the hypothesis.

Here we examine the relationshi pbetween rela-

tive species age and range size in the diverse shrub

genusPiper(Piperaceae) using publicly available

internal-transcribed spacer (ITS) sequences from

GenBank (www.ncbi.nlm.nih.gov). Most of these

sequences were originally published in Jaramillo

and Manos (2001) and Jaramillo and Callejas

(2004a,b). We chosePiperbecause its species

are prominent and important members of many

rainforest communities throughout the world

(Jaramillo and Manos 2001, Marquis 2004),

there is a reasonably large amount of species-

level informative genetic data available, and this

taxon is an ideal model system for the study of

ecology and evolution (Dyer and Palmer 2004).

We focused our analysis on neotropical species

because many sequences were available for these

species, the biogeography of neotropical species

has been studied (Marquis 2004, Quijano-Abril

et al. 2006), and the range sizes of many species

could be estimated using data from the Missouri

Botanical Garden’s online database, W^3 Tropicos

(http://mobot.mobot.org/W3T/Search/vast.html).

We used Bayesian inference to infer a phy-

logenetic tree, and then used this tree topol-

ogy to estimate relative divergence dates among

the species using the program BEAST v1.3

(Drummond and Rambaut 2007), which uses

a Bayesian relaxed-clock approach to divergence

time estimation (Drummondet al. 2006). For the

phylogenetic inference, we aligned 113 sequences