Yesterday was a little light on posts as I was in transit for most of the day to a lecture at NYU by Kevin Hunt of Indiana University called "The inferred forest home of the earliest hominins: Firm foundation or house of cards?" The talk was much more narrow in scope than what it might sound like from the title however, and there were good points and relatively weak points made throughout the presentation. Here I present a few thoughts from notes I took during the lecture and conversations with the Rutgers professors and grad students that also attended after the lecture;
When I hear the term "savanna" I think of an open grassland with very little tree cover, inhabited by lions, giraffes, and various grazing bovids, a woodland (40-80% tree cover but not a entirely closed canopy) and a forest (closed, interlocking canopy of large trees) being quite different. In the past, however, some researchers have played fast and loose with the terms savanna, grassland (which I take as being synonymous with savanna), woodland, and forest, the term "savanna" sometimes being used to describe everything from desert to forest. Variations like "open forest" (= woodland?) or "closed woodland" (= forest?) have further confused the issue, and Hunt was spot on in identifying that there's a "failure to communicate" about what certain faunal assemblages might indicate and paleoecology might indicate at a site. He primarily drew his examples from "Mio-Pliocene mammals from the Middle Awash, Ethiopia" by Haile-Selassie et al., noting that the fauna described in the paper (although with other data sources) appear to indicate a grassland setting while the paper says that it was a "wooded habitat." Was there a disconnect between the people who were doing the faunal analysis and the lead authors, or are we just not being rigorous enough with our terms?
I applaud the fact that Hunt wants to create a sort of standardized listing of habitat types and characteristics, but as some people in the audience pointed out during the Q&A the fossil record gives us no clue as to whether an area had between 40 and 80% canopy cover; whatever definitions are generated have to be consonant with the types of data retrieved from paleontological sites in order to make sense. Beyond that, when we say a habitat was a mosaic of several habitat types it might help to get an idea of the relationships of the habitats to each other. Was the forest bounded by woodland that transitioned into grassland? Do we have any idea of the proportions of each (was it mostly grassland with narrow riverine forests?). A little more resolution is needed, but this won't be accomplished via one paper and hopefully it'll kick off a bigger debate about the usage of terms.
All of the above is significant because it directly relates to our ancient relatives in the region, Ardipithecus ramidus being specifically mentioned here. The choice of Ardipithecus is a little problematic, however, as some see it as being more closely related to chimps and there's still Orrorin tugenensis from Kenya and Sahelanthropus tchadensis from Chad, all of which are somewhat vying for position of being a hominin ancestor most closely related to humans (Orrorin and Sahelanthropus were not mentioned during the talk). Furthermore, evolution works via populations; one species does not evolve towards one end across an entire range at the same rate at the same time, so even if we get a better resolution on one site or habitat for a hominin ancestor it doesn't mean we've answered the question of habitat for all members of that species everywhere. Modern chimpanzees, for example, live in a variety of habitats, the population at Mt. Assirik in Senegal living in a much more open habitat than the chimpanzees at Gombe or Tai, each having their own cultures and differences.
After the opening about the paleoecology of the Middle Awash, Hunt related the area to his research on chimpanzees in Uganda at Toro-Semliki, the apes living in riverine forests along a number of tributaries, venturing out into the woodlands by avoiding the open savanna as if it were poison. The behavior and habits of these chimpanzees were not directly related back to the Middle Awash for comparison, but it was at least inferred that they might have been living in a similar habitat as Ardipithecus and that there are two kinds of foods that the Semliki chimps eat present in the Middle Awash, one being tamarinds. Things got a little sticky here as Hunt mentioned that he favored a foraging origin for bipedalism (not unlike Jolly's "Seed Eaters" hypothesis) where early hominins or their ancestors stood bipedally to eat fruit and nuts on the ground and in trees, but I've never particularly liked this hypothesis as the answer for the origin of bipedalism. I don't doubt that early hominins and their direct ancestors stood up to gather food at least occasionally, but I don't see how such behavior provides enough of a selective force for bipdealism to fully develop (how would standing up to pick fruit translate into greater survivorship and/or great reproductive success?) and I think there is much more to the story. Anyway, Hunt mentioned that two fruiting bodies of plants present in Semliki today were present at early hominin sites in the past, attempting to make a connection. Perhaps early hominins and their ancestors at tamarinds, too, but there was a long list of other plants known in the fossil record as well as small vertebrae and invertebrate prey, so while there is this point or correlation I don't know if we can say more than "Maybe they ate this sort of food as well."
As I mentioned before, though, this is an analysis of just one area among many correlated to a modern-day habitat that may be influenced by the burning of grasslands by park rangers to help populations of kob that were previously devastated, so it's hard to make direct connections. Whatever criticisms or questions I might have, though, the main point of the lecture that I agreed with is a need for more rigorous definitions of ecological terms for use in the fossil record and more communication between paleoanthropologists and ecologists, botanists, zoologists, etc. in order to get better resolution on different sites.
- Log in to post comments
Speaking of grasslands, I've always wondered whether there's any intrinsic difference between savannas, veldts, prairies, steppes, and pampas, or if that's just what we call grasslands on different continents. I've heard of Mediterranean climates in North America, for example, or Alpine climates in South America, but I've never heard of an Asian prairie or an African steppe.
And if that's the case, is there justification for using any term other than "grassland" when referring to paleogeography?
HP; I'm not an expert on grasslands, but there seems to be little difference in these terms except in location/preference. When I think of "savanna" I think of Africa and when I hear "pampas" I think South America, and when I think of "prairie" I think of North America. Pampas, though, seems to have a more specific definition relating to location and rainfall in modern day South America, so to me that term shouldn't be used for anything else. Likewise, steppes seem to be somewhat climate limited and have a few more defining characteristics, although they are comparable to prairie.
The problem with paleogeography is that we can get an idea of climate and the kinds of plants and animals living in a place, but figuring out how many trees there were in a given area is impossible to determine and makes things difficult. Throw in the fact that species might get jumbled together a bit in waterways during floods and you can have a bit of a problem (perhaps savanna species get transported to a lake by a flooded river). I think "grassland" is the best general term that links the ones you mentioned, although under grassland there still is variation in terms of tall v. short grass, amount of tree cover, precipitation, climate, etc.
The fossil record does, indeed, give us a clue as to the amount of forest cover, through a number of proxies.
The definition of savanna is established, there is no need to change it. Indeed, I would say that making new categories would be a bad idea. However, using proxies to indicate relevant metrics would be good.
Savanna is a term usually reserved for tropics. In this context, a savanna is anything from regular grassland, even if seasonal, to woodland, as long as there is enough light to grow grass between the trees. It was never meant to be a fine-tuned measurement.
HP: Brian is right. These are just different terms.
However, only savanna is the correct term for the range of habitats we are talking about here. Pampas is both a habitat type and a geographical location. There are no African pampas, but the South American pampas is a savanna.
And yes, we don't want to use grasslands because most savannas are not grasslands.
Many, if not most, paleoanthropologists do not know this, I promise you that. Kevin is a smart guy, he probably does know it.
Thanks for chiming in, Greg; I was hoping you would. Overall I think Kevin did a very good job and made some good points, it was just a lot of information in a very small space. I'm looking forward to when he gets his paper out on ecological settings for sure.
Here in the Midwest the word savanna is used to refer to a very characteristic landscape which once occupied a substantial portion of the border areas of the Tallgrass Prairie Region. The trees, which (I'm guessing) shaded 20-30% of these areas, were mostly spreading oaks in the White Oak group, such as Burr Oak, Swamp White Oak, and White Oak. Tallgrass prairie grasses and forbs grew in the spaces between the trees.
I had Prof. Kevin Hunt as an undergraduate, and yes he does favor bipedalism as an adaptation to eating in the trees (i.e. bipedal in trees to reach food). But, he definitely does not believe that it is the only solution. He is, I would say, one of those people who believe that many factors create the end result. Bipedal for food is one reason, but overlooking the savanna and changing habitat may also be reasons. He had a long discussion on this in class. Anyway, he did his dissertation on this stuff, so he knows what he is talking about. He published a paper on the tree bipedalism hypothesis if you would like to read it: Hunt KD. 1996. The postural feeding hypothesis: an ecological model for the evolution of bipedalism. South African Journal of Science 92(2):77-90.
I tried to send you a message on this topic earlier today, but could not find it in my sent box afterwards. Can you let me know if you've received it?
Thanks Jason. I will definitely check out the paper, although what I wrote stemmed directly from what Hunt said during the lecture and he did not make mention of other factors that would have contributed to bipedalism outside of the postural feeding hypothesis.
help!!!!!!!!!!!!!!!!!!!!!!!!!!!!!1 i cant find any info on the savanna deserts!!!!!!!!!11