Parts 1, 2, and 3 of my Yachay talk dealt specifically with orchids. Here in Part 4 I discuss a more general philosophical question for conservation. Everyone agrees that one of our goals should be to conserve biodiversity. But what exactly do we mean by “biodiversity”? And how should we quantify biodiversity so we can assign priorities to ecosystems and conservation plans? Unfortunately time ran out at Yachay before I got to this part. But I want to write about it anyway….
The simplest measure of a community’s biodiversity is the number of species it contains. This measure is called “species richness” in the literature. That’s a useful starting point, though in rich tropical ecosystems most of the species are very rare, making it hard to accurately count them. The number of species found at a site often depends strongly on the sampling effort and the species relative abundances, so it is hard to make fair compare between sites. I’ve written earlier about a good solution to this problem.
Even if the number of species can be accurately measured, not all species are equal, so the total number of species present can be a poor measure of the value of a site. Threatened species have more conservation importance than widespread non-threatened ones. A good strategy would be to give separate counts of the number of species in each threat category, according to the IUCN classification or some other system. A count of local endemics would also be useful, regardless of their current threat level.
But even within a single threat category, not all species are equal. The loss of any species is tragic. But losing a distinctive plant like Amborella or Ginkgo biloba seems far more tragic than losing one of the ten pine species in the genus Pinus, or one of the many nearly identical species of asters. Most would agree that losing a phylogenetically isolated bird like a Hoatzin, which is in its own genus and even its own subfamily, would be more tragic than losing one of the juncos, which has many very close relatives.
The importance we assign to Ginkgo biloba, or the Hoatzin, comes from our sense that we should conserve “unique evolutionary history”. The more unique evolutionary history a species embodies, the more important it is for conservation. This evolutionary history is irreplaceable.
That’s a nice criterion because, with the advent of DNA analysis, we can quantify it more or less objectively. A time-calibrated phylogenetic tree of birds, for example, shows us exactly how long the Hoatzins have been evolving on their own isolated path (65 million years, since the extinction of the dinosaurs!). For any site and any group of organisms, we can construct the time-calibrated phylogenetic tree of the species present, and calculate how much unique evolutionary history the site hosts by adding up the branch lengths of that tree.
Dan Faith proposed this measure of conservation importance in 1992 [Faith D, 1992. Conservation evaluation and phylogenetic diversity. Biological Conservation 61: 1-10], and he called it “phylogenetic diversity”. He calculated it by adding up all the branches of the phylogenetic tree, beginning at the “root” of the tree, the point where all the branches came together at the most recent common ancestor of the group under study. See Figures 1 and 2.
This is a very useful starting point for incorporating evolutionary history into our conservation planning. However, it suffers from at least three problems in conservation applications. First, this phylogenetic diversity can depend very strongly on the vagaries of sampling; the location of the root of the tree can depend strongly on whether or not a particular species is detected in the sample. For example, suppose we are measuring the phylogenetic diversity of a New Guinea bird community. The Cassowary is a rare denizen of those forests, and that bird sits by itself on its own very deep branch in the New Guinea avian tree of life.
If the Cassowary is not detected in the sample, the root (most recent common ancestor) of the observed species will be very high up in the avian tree. But if the Cassowary is detected, the root will be pushed way back into the past, and the phylogenetic diversity will change not just by the amount represented in the Cassowary lineage, but also by an additional amount due to the lowering of the root of the rest of the avian lineages. See Fig. 3. This “double” sensitivity to sampling error is problematic.
The second problem is related to the first. The double sensitivity exaggerates the distinctiveness of the Cassowary; a new branch in the non-Cassowary group is being counted because the Cassowary appeared in the sample, but the evolutionary history represented by this branch is not from the Cassowary lineage. Adding the Cassowary adds 90 My to Faith’s phylogenetic diversity, but the Cassowary contributes only 50 My of that (Fig 3).
Third, and perhaps most important, this method of calculating the phylogenetic diversity of the group of interest leaves out an important part of the group’s evolutionary history, the part that precedes the common ancestor of today’s species and goes back to the time when the group diverged from the rest of earth’s organisms. See Fig. 4. This evolutionary history is unique to the group in question, not shared with any other group, and should be counted when we calculate the conservation importance of that group. In our hypothetical example, that branch contains an additional 50 My of unique evolutionary history embodied in this community (Fig. 4).
My coauthors Anne Chao, Chung-Huo Chiu, and I (Chao et al 2010) proposed to solve this problem by calculating phylogenetic diversity starting from the point of origin of the group in question, the point at which the study group diverged from the non-study groups. For the New Guinea birds, this would be the most recent common ancestor of avian birds and reptiles. See Fig 5.
This choice depends only on the study protocol, the decision (made in advance of executing the study) about what to include and exclude from the sample. It does not depend on the actual species sampled. If a study is aimed at birds, it will exclude moths, bats, and snakes that get into the mist nets. So if we want to measure all the unique evolutionary history contained in the focal group, birds, we should include the base of the bird tree, back to the point where it diverges from non-birds.
We might also want to divide this analysis according to threat category, or we might want to consider only the unique endemic species of a region. But these are details that depend on the goals of a study; there is no right or wrong choice about that.
So, suppose we want to use our suggested measure to find the amount of unique evolutionary history contained in the birds (or other focal group) of a site. This measure obliges us to construct a phylogenetic tree for the site’s species. Almost all phylogenies are reconstructed from neutral marker gene segments, which are not subject to natural selection. These trees may be time-calibrated, so that branch lengths are proportional to evolutionary time. All branch tips end at the same time, the present time. The technical term for this kind of flat-topped tree is “ultrametric”, meaning that the distance from the tree’s base to a branch tip is the same for every branch tip. Phylogenetic diversity calculated from a time-calibrated tree will be expressed in units of time. It can be calculated by multiplying the mean species richness of the tree (over the appropriate range of times from base to branch tips) multiplied by the length of the time interval between the base and the branch tips. A better term for it would be “phylogenetic work”, by analogy with physics.
Not all trees are ultrametric. Trees are often drawn with branch lengths proportional to the number of DNA base changes on the branch. The total number of base changes is not always the same in every terminal lineage, so the tops of these trees are usually uneven. Nevertheless we can still sum the lengths of the branches and obtain a measure of evolutionary work done to the focal community since the time it diverged from the non-focal groups.
It is now widely recognized that most evolutionary change, at the molecular level, is neutral or nearly so. Thus, evolutionary histories based on trees built from neutral alleles are likely to reflect the overall amount of molecular diversity and the amount of molecular innovation contained in the community. However, much of this variation is random noise that has no functional value.
For conservation it might make sense to value functional innovations over random selectively-neutral variations. Species that are under constantly-changing natural selection might be expected to evolve more innovations than species in a stable environment. To quantify this, we’d have to make phylogenetic trees whose branch lengths were proportional to changes in functional DNA. Not much work has been done along these lines yet. An alternative would be to develop a direct measure of functional diversity based on morphological, behavioral, or chemical differences between species. However, we rarely have enough detailed information about these aspects of species. We expect DNA-based phylogenies to be good proxies for this kind of diversity, and these are much easier to generate. Chao et al (2015) have developed a unified framework for functional, phylogenetic, and ordinary diversity, which should be helpful for this purpose.
All the measures of diversity I have just discussed ignore the abundances of the species. That’s fine for classical species-centric conservation; mere presence of a viable population of a species at a site may be enough for conservation of that species. But perhaps especially complex ecosystems are worth saving in and of themselves, regardless of whether the component species are threatened. Especially interesting are ecosystems where abundance is distributed evenly among many different phylogenetic lineages, instead of being concentrated in just a few. An ecosystem in which many diverse lineages are fully integrated and interacting with each other is very unusual and may have lots to teach us.
There are lots of diversity measures that do take into account abundances. Their use and abuse is one of the strangest stories in the recent history of science. Someday I’ll tell their story here, but meanwhile readers who want to know about them can read the following articles:
Jost, L. (2006) Entropy and diversity. Oikos 113: 363–375.
Jost, L. (2009) Mismeasuring biological diversity: Response to Hoffmann and Hoffmann (2008). Ecological Economics 68: 925–928.
Jost, L. (2007) Partitioning diversity into independent alpha and beta components. Ecology 88: 2427–2439.
Jost, L. (2008) GST and its relatives do not measure differentiation. Molecular Ecology 17: 4015–4026.
Jost, L., DeVries, P., Walla, T., Greeney, H., Chao, A. & Ricotta, C. (2010). Partitioning diversity for conservation analyses, Diversity and Distribution, 16, 65–76.
Chao, A., Chiu C.-H. and Jost, L. (2010). Phylogenetic diversity measures based on Hill numbers. Philosophical Transactions of the Royal Society B 365: 3599-3609.
Chao, A., Chiu, C. H., & Jost, L. (2014). Unifying Species Diversity, Phylogenetic Diversity, Functional Diversity, and Related Similarity and Differentiation Measures Through Hill Numbers. Annual Review of Ecology, Evolution, and Systematics, 45, 297-324.
Lou Jost, www.loujost.com