Research focus: Will Brightly, PhD candidate in the Strömberg Lab I am interested in the ecological impacts of the Cenozoic expansion of grassland ecosystems, particularly from the perspective of the most important organisms in those ecosystems: the grasses themselves. One of the projects that I have been working on recently focuses on how seed dispersal may have changed during this transition. To investigate this, I am looking at the rich fossil record of silicified grass anthoecia (these are the bracts that enclose the grass fruit) preserved in the Oligocene-Miocene of Kansas and Nebraska. Because the expansion of grassland ecosystems would have been associated with a number of changes relevant to seed dispersal (e.g., increased abundance of grazers, increased access to wind for plants of smaller stature), we suspect that some of the morphological variation we see in this record may be related to changes in seed dispersal strategy. But before we can evaluate this, we have to figure out how we can use the morphology of anthoecia in modern grasses to infer dispersal mode. One of the ways that we are approaching this question is to quantify traits related to different modes of seed dispersal. One of the seed dispersal modes we are specifically looking at is adhesive dispersal. Here, we can exploit the premise that seeds (or the structure containing them) are more likely to become attached to passing animals if they have lots of hooks, hairs, and other protrusions. This is essentially the surface roughness of the object, which we can easily quantify if we have some good photos of our fossils (and modern references). We have therefore been spending lots of time at the Burke Museum over the past several months, photographing specimens from the Burke Herbarium and other herbaria and using the data we collect from these images to construct our modern reference collection. Among other things, this consists of grasses with well established seed dispersal syndromes and the nearest living relatives of the fossil taxa we are working with. So far, we have collected data for 120 individuals from around 50 species (though we are a long way from done). Very soon we will get started on collecting data for our fossil taxa, which is when the real fun begins! Finally, I wanted to quickly run through two examples to illustrate how we’re collecting our data, and take a look at two ends of the adhesive dispersal spectrum. First let’s take a look at Cenchrus palmeri, a species that is characteristically adhesively dispersed. This species (and its relatives) produces burrs that are very good at sticking places they’re not always wanted. Indeed, you may have had some experience with grasses in this genus in the past. We’ll compare these burrs to the florets (these are the anthoecia plus the reproductive structures they enclose) of Panicum miliaceum, a species that is commonly grown for birdseed (and occasionally humanseed). Just by intuition we can predict which of these two species is more likely to have seeds adhesively dispersed. However, not all taxa are this cut and dry, and it is useful to have a quantitative way to compare taxa that fall between these two extremes. We start by taking photos of the relevant unit of seed dispersal in three different orientations (a dorsal/ventral view, a lateral view, and a top view). We then use a thresholding tool to outline each and record the perimeter of the object. Next, the object outline is divided by the perimeter of the convex hull of the object, which provides a measure of its roughness. As surface roughness decreases, this metric will increase to a maximum value of one, and as surface roughness increases, values will approach zero. We then take an average of our three different views (weighted based upon the surface area presented in each) to obtain our overall estimate of surface roughness for the object. Applying this method to our two example taxa, we get a value of 0.896 for Panicum miliaceum and a value of 0.263 for Cenchrus palmeri. This is exactly what we expected based upon what we already know, and can observe about these two grasses (remembering that as roughness increases we approach a value of zero). Now let’s see what we can find out about the fossils!
0 Comments
Leave a Reply. |
AboutWelcome to the blog page of the Strömberg Lab! Archives |