Gypsophily Research

The Origins and Biogeography of Gypsophily in the Chihuahuan Desert

Edaphic endemism, or the restriction of a plant species to a particular geological substrate or soil type, is a common phenomenon in plants, particularly in the drier regions of the world. In fact, on a global scale, substrate ranks close behind moisture availability and temperature in determining the geographic distribution of plant species.

Approximate geographic extent of the Chihuahuan Desert (in gray).

One of the largest edaphically restricted floras in North America is found in the Chihuahuan Desert region of north central Mexico and the southwestern United States. The Chihuahuan Desert is the largest desert in North America, and in spite of its incredible biodiversity, is also the least studied of the major North American deserts. While much of the region is underlain by limestone, exposed gypsum deposits are also a relatively common feature of the Chihuahuan Desert. These gypsum deposits are distributed in a discontinuous, island-like fashion throughout the Chihuahuan Desert region. Gypsum, or hydrated calcium sulfate (CaSO4·2H2O; this is the major ingredient in plaster of Paris and wallboard), is a difficult substrate for plants to germinate and survive on because it typically forms a hard crust when dry, erodes quickly when wet, and is relatively low in available nutrients. Despite the inhospitability of gypsum, a large and diverse group of Chihuahuan Desert plants are found on no other substrate. Examples of these gypsophilic (“gypsum-loving”) plants include members of the genera Acleisanthes and Anulocaulis (Nyctaginaceae), Nerisyrenia (Brassicaceae), Nama and Tiquilia (Boraginaceae), Drymaria (Caryophyllaceae), Fouquieria (Fouquieriaceae), Argemone (Papaveraceae), Mentzelia (Loasaceae), Calylophus (Onagraceae), Gaillardia, Sartwellia, Haploësthes, Xanthisma, Dicranocarpus, Marshalljohnstonia, and Strotheria (Asteraceae), and Sporobolus (Poaceae). Scroll down for images of some of these gypsophiles.

Above: Image of gypsum showing surface crust that often forms.

Above: gypsum dunes at Cuatro Ciénegas, Coahuila, México. The plants forming low mounds in the foreground are Tiquilia turneri, a gypsophilic species.

Above: White Mesa, northwest of Albuquerque, NM. The crest of the mesa is composed of a thick bed of pure gypsum.

Above: pure gypsum rocks at Chupadera Mesa in central NM. The yellow-flowered plant at right is Sartwellia flaveriae, a gypsophile.

Age of Chihuahuan Desert Gypsophily

Aside from the large number and diversity of species, two other attributes make the Chihuahuan Desert gypsophilic flora interesting from a systematic perspective because they suggest that it may be a relatively old assemblage. First, several plant genera in the Chihuahuan Desert have multiple gypsophilic species (examples include Argemone, Drymaria, Gaillardia, Nama, and Tiquilia), and a handful of genera (such as Nerisyrenia and Sartwellia) are composed almost entirely of gypsophiles. Within most of these groups, the physical similarities shared among the individual gypsophilic species suggest that these species form a monophyletic group; i.e. these species appear to be each others’ closest relatives. Should this prove to be true, it would indicate that speciation (the formation of new species) occurred after these groups became restricted to gypsum. Second, the same gypsophilic species are often found growing together across much of the Chihuahuan Desert, even though the gypsum deposits these plants inhabit may be isolated from one another by dozens (even hundreds) of kilometers of non-gypsum substrates. Since the only way for a gypsophilic plant to move from one isolated deposit to another is via chance long-distance dispersal, it makes sense that these gypsophiles must be relatively old to account for the time it must have taken for so many of them to have spread so widely.

How can we prove that the gypsophilic plant species of the Chihuahuan Desert are as old as they seem? Unfortunately, fossils do not form well in arid regions, and thus we cannot rely upon a fossil record to determine the age of these plants directly. Instead, we must take a molecular phylogenetic approach. By reconstructing the evolutionary relationships within each of these gypsophilic groups using of DNA sequence data, we can indirectly estimate the age of gypsophilic group using molecular dating techniques. Roughly speaking, these techniques involve assessing the amount of DNA sequence evolution that has occurred along an evolutionary branch within a phylogenetic tree, and then converting this quantity to an age through the use of a known time calibration point. Although this method of inferring the ages of lineages is not without limitations, it has been shown to provide useful age estimates in many instances.

Images of gypsophiles:

Acleisanthes purpusiana, central Coahuila	Gaillardia multiceps, west Texas

Calylophus filifolius, west Texas	Haploësthes gregii var. texana; this species is not a strict gypsophile, but most of its close relatives are.

Nerisyrenia linearifolia, southern New Mexico	Anulocaulis leiosolenus var. lasianthus, west Texas

Fouquieria shrevei, central Coahuila	Sporobolus nealleyi (the ring-forming grass in this image), southern NM

In addition to inferring their antiquity, a molecular phylogenetic approach can also inform us about other elements of the evolution of gypsum endemism. For example, all of the genera that contain multiple gypsophiles within the Chihuahuan Desert also contain species that grow both on and off gypsum (so-called “gypsovags”). Where do the gypsophilic species within each group fall out with respect to these gypsovags? Do the gypsophilic species form a clade (a monophyletic of species) embedded within a group of gypsovagic species (implying a single origin of gypsophily—see Scenario 1 below), or are they interdigitated with the gypsovags (implying multiple origins of gypsophily—see Scenario 2 below)? The former result would suggest that the transition to gypsophily is a relatively rare occurrence, perhaps reflecting the relatively difficult conditions plants must endure on gypsum, while the latter result would suggest that gypsophily is less difficult to attain. Again, we might also expect to see the former result if the gypsum flora is a relatively old one.

Finally, we can use molecular phylogenetic techniques to evaluate genetic diversity within a gypsophilic species. This becomes especially important for the many widespread gypsophilic species, because we might reasonably expect them to display relatively high levels of interpopulational genetic variation. This hypothesis makes sense because of the isolated nature of surface gypsum exposures in the Chihuahuan Desert; the low likelihood of gypsophile dispersal between these deposits would promote genetic isolation among gypsophilic populations. We can test this prediction by searching for DNA sequence variation among individuals from different populations of the same gypsophilic species.

My earlier phylogenetic research on the plant group Tiquilia subg. Eddya, which contains two narrowly distributed gypsophilic species, one very widespread gypsophilic species, and seven species of gypsovags, indicated that gypsophily arose twice and is relatively old (early Pliocene, ~5 million years old). This age corresponds closely with the first onset of relatively dry conditions in the American Southwest that permitted the exposure of gypsum at the surface over broad areas. Consequently, it would appear that almost as soon as gypsum exposures became available for desert plants to occupy, some plants nearly immediately became restricted to gypsum. Also exciting was the discovery of a tremendous amount of DNA sequence variation in the widespread gypsophile Tiquilia hispidissima (right). Moreover, the genetic diversity in this species was strongly correlated with geography, precisely as we would expect in a widespread gypsum endemic. Perhaps the most exciting find was that the geographic distribution of genetic diversity (also called phylogeographic diversity) in Tiquilia hispidissima corresponds to the geographic distributions of closely related gypsophilic species in several other Chihuahuan Desert plant groups. We do not have the detailed phylogenies of most of these other groups that would be necessary to confirm whether the geographic pattern of genetic diversity in these groups is similar to that observed in Tiquilia hispidissima. A common phylogeographic pattern among these groups would suggest the existence of several historical centers of gypsum endemism around the Chihuahuan Desert that have played important roles in generating genetic diversity in these plants.

Current and Future Research

I am in the preliminary stages of extending my former research on gypsophily in Tiquilia to several other major gypsophilic lineages in the Chihuahuan Desert. In particular, I am interested in examining whether the phylogeographic pattern observed in Tiquilia hispidissima represents a more general pattern among widespread Chihuahuan Desert gypsophiles. I am also interested in using molecular dating techniques to determine whether other gypsophilic groups first became restricted to gypsum in the early Pliocene, similar to Tiquilia. Finally, I want to investigate whether multiple phylogenetic origins of gypsophily is a general phenomenon in Chihuahuan Desert gypsophilic plant groups.

I also intend to explore the ecological origins of gypsophily in the Chihuahuan Desert. What kinds of adaptations allow for Chihuahuan Desert gypsophiles to thrive on gypsum? Do these plants have a physiological requirement for gypsum, or is there some physical characteristic of gypsum substrates that normally prevents germination and establishment, but that gypsophiles have overcome? Do gypsovags respond differently in this regard? To answer these questions will require growing and experimenting with plants on different types of substrates, both in the field and in the greenhouse.

Top

last updated November 30, 2007