Terry 2010 paper
In this article, Terry’s primary goal is to determine
whether utilizing fossilized remains (in this case specifically small-mammal
prey caught by owls) provide a valid and accurate representation of living
communities and ecosystems over time. Focusing mainly on the predation of owls
on small rodents in the Two Ledges Chamber cave system in Nevada, Terry
compares the long term concentrated small-mammal death assemblages (post-mortem
rodents eaten and regurgitated into pellets, thus preserved) with data from
modern trapping surveys to determine the
correlation in the live-dead agreement. Terry’s primary goals are: 1) to
evaluate the ecological fidelity of the modern small-mammal community of Two
Ledges Chamber. 2) determine the effectiveness and accuracy of the technique
over time (from months to centuries). 3) Finally, to assess how spatial
averaging affects the landscape-picture as reconstructed from the
skeletal/fossilized remains, and determine to what degree have modern
communities/ecosystems been altered.
For her study Terry found a cave that consistently had owl
pellets which contain local small mammal skeleton remains in them. Using this
and the fact that the remains weren’t very obstructed by humans she was able to
pull a source of fossils, surficial skeletal assembles and modern skeletons
from today to compare to a poll from traps that were set to capture modern
small mammals that were statistically consistent in the owls diet. Using the mammal data she
then spread out the range to 5 locations to observe where the population comes
from as well as abundance (Figure 6). To
analyze her data she compared the percentage of living species to the dead
assemblage and compared the dead species found in the live community to
determine the live-dead agreement of the species. To analyze her statistical
data she used Spearman rho and Bray-Curtis similarity index to determine that
the agreement between all samples were high. This means that the fidelity of
the small mammal community has remained similar throughout time, thus giving
paleontologists and neonatologists new tools and a valid model to quantify
“live-dead agreements” of terrestrial communities/ecosystems.
The greatest draw back of the study stemmed from the spatial
arrangement; while it was found that the small mammals that were pulled from
the 5 different locations were consistently similar over time, location did
play a role as the pellets were mainly focused around the cliffs and cheatgrass
showing the highest live-dead agreement. The cliff showed the highest agreement
meaning it was hunted the most, and thus accuracy of the study might decrease
in areas not suitable for hunting. How could Terry’s techniques/approach be
changed to accurately assess an environment with low predation or one where the
pellets are not deposited? Also the environment created by the Two Ledges Caves
is very ideal for preservation of fossil/skeletal remains. How would Terry’s
proposed model fare in different environments and would it still be valid for
understanding the taphonomy of terrestrial systems?
In an environment with low predation or one where the pellets are not deposited, Terry’s techniques would need to be altered such that remains of natural deaths are found. This is incredibly difficult since these results focused on rodent species, which commonly use burrows for shelter. As for other locations: her model would not be valid. If this model were to be used in other locations, it would be used as an introductory model and then expanded upon or rejected. Every ecosystem is different, and run through their own routes; as such every location would need its own system to understand taphonomy.
ReplyDeleteTheoretically, I believe Terry’s proposed model would work in any environment; however, the difficulty would be in accounting for: (1) the habitat of the species of interest, (2) differential preservation within the environment, and (3) finding a ‘dead’ assemblage to correlate to a ‘live’ assemblage.
ReplyDeleteFor instance in a tropical area, if you are interested in an animal that frequents rockshelters such as bats or armadillos, you are likely to be able to find both a ‘live’ and ‘dead’ assemblage. The difficulty would then be if you are looking for a species that does not frequent rockshelters, such as a peccary. Because of different preservation levels within and outside of the rockshelter, you would have a harder time finding a ‘dead’ assemblage, that had not been affected by humans, of peccaries than that of bats or armadillos.
To relate back to the first question posed by Derek Cravens, in an environment with low predation (or if you are looking at a herbivore) or one where pellets are not deposited (ie you are not looking at owls), you would need to adjust where you are sampling, such as frequenting places where that animal hunts or forages the most. The difficulty would arise in obtaining a high number of surface skeletons within a defined area. I suspect this may involve numerous forays into the area to search and/or help from people living or working in the area if possible.
Having now read both Terry’s and Miller’s studies, I can appreciate the importance of studying death assemblages to reconstruct historical baselines and assess ecological fidelity of terrestrial species. While Miller’s study looked at large ungulates, Terry’s study provides important insight on small mammals, specifically trying to assess the fact that these are often preyed upon. I found Terry’s study to be interesting due to the complexity of factors involved such as predation, capture of ‘live data” through trapping surveys, and data gathering techniques via cross-temporal owl pellet collection. I found her techniques to be functional and innovative considering that pellets tend to conserve death assemblages of small mammals incredibly well and agree with the author that this technique should be valued. However, how would a study like this be done for other small mammals that are not necessarily preyed upon by animals that produce pellets? How can we study small mammals in other ways to reconstruct historical baselines in ecosystems that have been influenced by humans?
ReplyDeleteWhen reading Terry's paper, she speaks about the predator/trapping bias. She concludes that the owls and traps catch and collect the same species, but in different abundances. I wonder if there could be a better way to have similar species abundances in trapping?
ReplyDeleteAlso, on a side note about predators, could predators cause a mass extinction or are environmental catastrophes the only cause of mass extinctions?
In Terry's paper she goes into depth on how she collection the data, such as trapping of live data and owl pellets from predation of these small mammals. Although these methods may be effective in collecting one set of data based on predation what about other symbiotic relationships such as competition or parasitism to collect more data about natural death assemblages? would these process work just as effectively in collecting data?
ReplyDeleteGiven the integral nature of both predation and death assemblages in the scope of Terry’s research, I find it difficult to contemplate how her methods could be adapted to an area where both predation and death assemblage deposits are low. It would seem, to me, that such conditions would require a departure from this method, to those that aren’t dependent upon death assemblages or predation behaviors. Further, I should think that Terry’s work could be adapted to different environments provided there is the ability to collect relevant live species data, track a target predator species, and assess the related death assemblages (as done in the current research study). As I was reading the paper, the idea of predator preferences (in terms of prey type) playing a role in death assemblage composition, stuck out to me. If this work were to be carried out elsewhere, I think that this issue would have to be further addressed in the scope of the environment of interest, in order to reduce bias.
ReplyDeleteStudies like these seem to be limited by the type of prey that these raptors (or other predators) can catch. It would be expected that more exclusively fossorial and sub-surficial taxa would not show up in these pellets as there is just a less likelihood of them being consumed by the raptors. I think that in order to best reconstruct the ecology of these ecosystems, we need to have both a natural death-bed assemblage (Konservat-Lagerstätten) in addition to the predatorily conceived one (Konzentrat-Lagerstätten) analyzed in the paper.
ReplyDeleteSeveral things concern me about this study.
ReplyDelete1) I don't know one way or the other, but I doubt suitable owl pellet sites like these are very common, even by paleontological standards. Maybe they're all over the place though, IDK.
2) The oldest dated bone in the study was only ~9020 years old, indicating that the record at this site is not very long.
3) As mentioned by some previous commenters, an owl pellet based record will show only a very narrow selection of the total ecosystem (namely small mammals). Indeed, the predatory habits of owls, however opportunistic they may be, may introduce a sampling bias against burrowing mammals which spend little time above ground.
3) Terry describes the "live-dead agreement" between modern and fossil records as "high", though there are some discrepancies. I wonder, how does Terry explain these discrepancies? Are they the result of differential preservation of fossils? Are they the result of bias introduced by owl-based prey preferences? Are they the result of actual changes in the ecosystem over time?
Terry tries to parse out some of these potential sources of confusion by comparing the results of the trapping surveys to the modern owl pellets, but this does not address either differential preservation and fossilization of smaller/larger mammal bones, or the possibility of changes in owl feeding behavior over time, or the possibility that the opportunistic Barn Owls which are currently dominant at the site may have previously been less common, or shared the site with other, now absent, owl species with different feeding habits.
Most importantly, if we looked at another system where the "live-dead agreement" between modern and fossil records was not "high", how could we tell the difference between preservation/predation bias VS actual ecosystem change over time? If we wanted to quantify how an ecosystem changed over time using this method, or even how a specific species of rodent varied over time, we would have to worry about the apparent 'change' we saw being nothing more than an artefact of some predation-introduced error. This would be especially concerning for any study that spanned a time period significantly longer than ~9020 years, since more ecological changes (both of the prey items being measured as well as the predators which serve to concentrate their remains), more bone dissolution, and more bioturbation may have occurred.
In short, I would be reluctant to draw any overarching conclusions from this method alone. It seems to me that it is best suited as a supplemental analysis. Of course, drawing your conclusions from multiple studies is usually preferable to using one avenue alone, so I'm not really saying anything too special. Concordance is good!
If the model was expanded to encompass areas of low predation and areas where no pellets are found her system would not be able to accurately predict populations or communities living in the past. She would need to broaden her search for remains and evidence of owl predation as well as new ways to find evidence of rodent populations where owls aren't actively hunting due to varies reasons which could be accessibility or climate. This would be difficult to do because undisturbed remains of owl meals are very well preserved and tell a lot more information then scattered bones in a forest. Humans have completely changed the natural environment and it would be hard for this model to account for human variables.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteIt seems that this method could prove fruitful for ecological research within the last few thousand years, provided that a few caveats be kept in mind. Owls, like many predators, will show preference in their prey items. Also, like many predators, they have the capacity to become more proficient in capturing certain prey items, over the generations that were once less preferred because of difficulties in their capture (meaning that an increase in these bones may be due to greater capture efficiency rather than increase in population). The method also favors smaller mammals, that can be more easily swallowed whole to yield more intact bones. Great Horned Owls prey upon larger prey items, such as skunks, and could therefore be more difficult to identify in their pellets. Lastly, this method depends upon a good source of owl pellets, deposited over the years,with which to work with. This may prove difficult in areas that lack good owl perches and nesting sites such as in open prairies deserts.
ReplyDeleteTerry's paper showed that there is a "live-dead agreement" in the rodent population that spans from three hundred years ago to modern day. Of course there could be problems with this data like the things we discussed in class, such as owls only hunting at night or that they need perches to spot their prey. However, I think that Terry corrected for these problems the best that she could. Overall I think the death assemblage method is a very useful tool in comparing past and present animal populations.
ReplyDeleteTerry's study does have benefit and summat our knowledge of the small mammal prey of owl,over the couple of a few thousand years. As for the technique used it has too many variable factors such as predation, preservation of fossils and locality. Does this study only work on owls or would it do other predatory organisms? The most important point is what if the organism chosen to study is not a pallet producing predator?
ReplyDelete