Mount Claremont, Australia
Mount Claremont, Australia

Time filter

Source Type

No statistical methods were used to predetermine sample size. The data for each analysis consist of a binary presence–absence matrix in which each row is a taxon and each column is a sample. The entries represent the presence (1) or absence (0) of a particular taxon in a particular sample. Within this matrix, each of the S(S − 1)/2 unique species pairs is tested and classified as random, aggregated, or segregated. The tests were performed with the PAIRS version 1.0 software application15, 31. The methodology is described fully in ref. 2 and is briefly described here. The analysis begins by calculating a scaled C score32: C  = (R  – D)(R  – D)/R R , where C is the C score for species pair i and j, R is the row total (the number of species occurrences) for species i, R is the row total for species j, and D is the number of shared sites in which both species are present. For each species pair, this index ranges from 0.0 (aggregation: maximal co-occurrence of both species) to 1.0 (segregation: minimal co-occurrence of both species). PAIRS calculates the C score for each pair of species and assigns it to a histogram bin. There are 20 bins that range from 0 to 1 in 0.05 intervals, plus a bin for 0.0 (perfectly aggregated pairs) and a bin for 1.0 (perfectly segregated pairs). We next estimate the P value associated with each species pair by a randomization test. The data matrix is first randomized by reshuffling all matrix elements, with the restriction that the row and column sums of the original matrix are preserved. This ‘fixed-fixed’ algorithm has been subject to extensive benchmark testing with artificial random and structured matrices2, 33, 34. Compared with a variety of other null model algorithms, the fixed-fixed algorithm most effectively screens against type I errors (incorrectly rejecting the null hypothesis for a random matrix), but is somewhat conservative33. An alternative algorithm ‘fixed-equiprobable’ retains row sums (species occurrence frequencies), but allows column totals (species richness per site) to vary freely. The fixed-equiprobable algorithm also has good statistical properties, and is appropriate for modern data sets in which sampling effort has been standardized, such as quadrat samples of fixed area. However, this algorithm is not appropriate for fossil data because the number of species detected per site in fossil assemblages is determined primarily by sampling effort of the collector and by site-specific taphonomic biases in preservation. For these reasons, we have used only the fixed-fixed model, both for the analysis of fossil assemblages and for comparison with modern assemblages. Details of the randomization are discussed further in refs 2, 35. Using 1,000 randomizations, we create a simple P value (two-tailed test) for each species pair, which leads to a classification of each species pair as aggregated, random, or segregated. However, with a total of S(S − 1)/2 unique pairs in a matrix of S species, retaining all of the significant pairs (P < 0.05) would generate a potentially large number of false positive results. This problem has frequently arisen in the analysis of micro-arrays, genomic surveys, and other examples of ‘big data’36. The PAIRS analysis relies on an empirical Bayes approach by creating a histogram of C score values based on the pairs generated in each null assemblage. To screen out false positives, we calculated the average number of species pairs in each bin of the histogram. Next, we determined the observed number of pairs from the empirical assemblage in each bin, ordered by their P values from the simulation. We retained only those pairs that were above the mean number for each bin and were statistically significant on the basis of the simple P value criterion. This double screen effectively eliminates many of the false-positives that can arise in random data sets2. A Loess smoothing line was created with the stat_smooth function in the R package ggplot2 version 1.0.0 (ref. 37) using default parameters. For Loess fitting, the fit at point x is made using points in the neighbourhood of x (closest 75% of total points), with tricubic weighting (proportional to (1 − (distance/maximum distance)3)3). Points were additionally weighted by the number of sites in each matrix and 95% confidence intervals were generated using a t-based approximation. To examine the how climate variability impacts the percentage of aggregated species pairs, we used climate proxy data from ice17 and deep sea cores16, which collectively encompass the past 65 Myr of the assembled data sets. The European Project for Ice Coring in Antarctica (EPICA) data were used preferentially when there was temporal overlap between proxy data sets. Climate data were mean centred and standardized before pooling into a single data time series. We then sampled the climate data across the ‘temporal extents’ (Extended Data Table 1) of the individual Evolution of Terrestrial Ecosystems Program (ETE) data sets to test if there were relationships between the percentage of aggregated species pairs and climate variability. Climate variability was calculated in two ways: (1) as the standard deviation of climate within the temporal extent of each data set and (2) as the standard deviation of the first differences (changes in climate from available time-step to time-step within the temporal extent of a data set) of climate. We used standard deviation because it helps address issues with changes in data density over time. Estimated rates of change are sensitive to the time span over which they are measured and more closely spaced data would shift apparent rates of change. Approaches using standard deviation are less sensitive to this issue. We also compared climate variability with age (years before present) of ETE data sets to test for Phanerozoic-scale trends in climate variability sampled by ETE data sets. We used a maximum likelihood approach, available in the R package ‘segmented’ version 1.1, to estimate the breakpoint time at which the sharper decline in aggregated species pairs began. This analysis used an initial linear model of the proportion of aggregated pairs as a function of community age (log of years before present) to generate a best-fitting number of breakpoints, with separate regression lines fit to each segment. A bootstrap of 1,000 replicates was used to estimate uncertainty in the model parameters (including uncertainty in the time of the breakpoint). Collection modes. We thought that differences in the way fossil and modern data are collected might be responsible for the observed difference in the relative proportions of aggregated versus segregated species pairs in modern communities2, 10 and fossil communities. There are two reasons why collection modes are not likely to be responsible for this difference. First, fossil collections are heterogenous by nature. Different collecting methods are used for different taxonomic groups (for example, bulk sampling, surface sampling, cores). Moreover, even within a taxonomic group, the type of depositional environment imposes different kinds of bias (for example, cave sites versus open pits for Pleistocene mammals). Second, we see a switch from species pairs that are dominated by aggregations to those dominated by segregations in our data sets that span the Pleistocene–Holocene transition (Extended Data Fig. 4 and Extended Data Table 1). In particular, mammal assemblages show a switch from >50% aggregations in the Pleistocene to <50% aggregations in the Holocene. The data encompassing this switch are all fossil localities and there are similar biases in both time slices. Although there is variation in the pollen assemblages, a weighted regression that takes into account the sampling in each time slice shows a significant decrease through time (P = 0.04, R2 = 0.15). This trend of increasing percentage of segregated pairs begins approximately 14,000 years ago and continues across the Holocene with the switch occurring in the final 1,000 year time slice20. The fact that these data were all collected using the same sampling techniques suggests that sampling cannot account for this pattern. Issues of scale. It is generally assumed that fossil data are biased. Although the type of bias is not the same for all taxonomic groups, most fossil assemblages contain some degree of temporal or spatial averaging38. That is, they represent accumulations of species that can occur over hundreds or thousands of years and may mix species that did not exist at the locality at the same time39. The fossil data sets in this analysis include assemblages that range from no time-averaging (for example, fossil leaves preserved in volcanic event beds) to those that are time-averaged over thousands or hundreds of thousands of years (for example, some mammal assemblages). In addition, some data sets could not be resolved to time bins of less than a million years. Spatial averaging is less of an issue in these data sets, but individual samples are drawn from areas with diameters ranging from a few metres to more than 300 km (Supplementary Table 1). If issues of scale are contributing to the pattern found here, there should be a relationship between the proportion of significant pairs that are aggregated and the spatial or temporal scale of the data. We evaluated this by estimating the spatial or temporal grain and extent of each data set included in the analyses (Extended Data Table 1) and determining if there was a significant relationship with the percentage of aggregations. The spatial grain is the estimated radius of collection area over which fossil specimens would have been transported to the depositional environment in a typical locality. The temporal grain is the typical amount of time-averaging of localities in a data set. Spatial extent is the longest linear distance between any two sites in a data set and temporal extent is the duration from the oldest to youngest locality in a data set. We found no relationship between the scale of the data sets and the proportion of significant pairs that were aggregated versus segregated (Fig. 2 and Extended Data Fig. 5). Regression analyses were not significant and explained very little of the variation in the data (Extended Data Fig. 6). The pattern of segregated versus aggregated pairs was not different in fossil versus modern assemblages because of biases related to the scale of fossil data. Taphonomic bias. How can taphonomy and palaeoenvironment affect species frequencies (richness) and spatial representation? The fossil record contains buried assemblages of species that were derived from living communities at different times in the past. Species representation (presence or absence) in individual fossil assemblages is a critical attribute of our data sets, therefore we need to consider how this variable might be biased relative to original associations of species in communities. Taphonomic processes operate during the transition of dead remains into preserved samples and thus control the biological information that passes from the living community into the fossil record39. These processes act as filters that can alter species representation in fossil samples in a variety of ways: (1) selective preservation of organisms with particular body compositions and sizes, for example organisms with and without mineralized skeletons, larger versus smaller individuals; (2) variable preservation of organisms depending on their population abundance, spatial distribution and life habits, for example aquatic versus terrestrial; (3) post-mortem or depositional mixing of species that did not live together (time-averaging), or separation of species that did (selective transport or destruction). Additionally, some types of environment are better represented in the depositional record than others, such as wetlands versus dry land surfaces. All of these add up to potential biases that could affect biological signals and the proportions of random versus significant species pairs, or the proportions of segregated versus aggregated pairs, in our analyses. However, the particular null model algorithm used effectively controls for major sources of taphonomic bias in the data set. This ‘fixed-fixed’ algorithm33 creates null assemblages that have the same species richness per sample, and the same number of occurrences per species, as the original data. Thus, if there are preservation biases that generate heterogeneity in the total number of fossil species per sample, or biases in the number of specimens per species, these are effectively controlled for in the analysis. Significant patterns of species aggregation are those measured beyond the effects of sampling heterogeneity in the occurrences of species or the number of species per sample. Similar sampling effects are controlled for in the modern data, which can also exhibit variation in the commonness or rarity of species and in the number of species per sample. Taxonomic resolution of the data. Fossils are not always resolvable to the species level and are frequently analysed at the genus level. This may have the effect of increasing geographical ranges and overlap between taxa, and may contribute to the dominance of aggregated pairs found in this study. To test whether this was the case, we analysed 18 of the data sets at the species and genus level (16 mammal and 2 plant data sets). If taxonomic resolution is driving the pattern, we expect to see an increase in the proportion of aggregated pairs when species are lumped into genera. We found that six of the data sets showed the expected increase. However, nine showed a decrease and three showed no change (Extended Data Table 2). Interestingly, one of the modern data sets on small mammals from the Great Basin had genetic information that indicated that some were cryptic species. When the analysis was re-run with the cryptic species identified, there was an increase in the proportion of significantly aggregated pairs (from 50% to 61%). This is in the opposite direction that we would expect if lumping species into genera artificially increased aggregated pairs. Taken together, these results suggest that the differences between species associations over the past 300 Myr and the present are not driven by the taxonomic resolution of fossil data. Sampling of abundant and rare species in fossil and modern data. The results of null model analyses of abundance versus presence–absence data are compared in ref. 10. The two kinds of analysis give qualitatively comparable results, although the abundance analyses are somewhat more powerful in detecting non-randomness. It is generally assumed that fossil deposits miss the rarest species in a community because preservation potential increases with abundance; more individuals means more opportunities for fossilization events. If rare species are more likely to form segregated pairs, we would expect to see more segregations in modern data sets because they should sample more of the rare species than comparable fossil data sets. Within fossil data sets, we would expect to see more segregated pairs in data sets with better sampling and more rare species. We evaluated this using a data visualization technique. We present the results of our analyses as a series of pairwise species by species matrices and order species by occupancy (see Supplementary Information: data sets). Occupancy decreases as one moves to the right on the x axis and up on the y axis. Species with the highest occupancy are close to the origin. The pairwise associations are colour-coded: grey for random pairs, blue for aggregated pairs, and red for segregated pairs. If increasing sampling of rare species is responsible for the pattern we document, then we would expect to see a preponderance of red, segregated pairs in the upper, right-hand corner of the species by species matrices. In particular, this should show up in data sets with better sampling and those that encompass the shift from more aggregated to more segregated pairs (for example, Pleistocene–Holocene mammals and pollen, modern mammals in Kenya, and modern plants in Wisconsin). This is not the pattern that we see. In fact, we find that segregated pairs tend to form with species of intermediate occupancy and that aggregated pairs form both with common species and with rare species. Differences in the sampling of rare species between fossil and modern data sets cannot account for the shift in species associations over time.


News Article | February 15, 2017
Site: www.eurekalert.org

Conservationists need to adopt a critical shift in thinking to keep the Earth's ecosystems diverse and useful in an increasingly "unnatural" world. That was among the conclusions of conservationists from every continent but Antarctica who gathered at the University of California, Berkeley in September 2015 to discuss the future of conservation. The meeting included a diverse mix of countries and of specialists, including ecologists, conservation biologists, paleobiologists, geologists, lawyers, policymakers and writers. Their discussions, summarized and published in Science on Feb. 9, recommend a more vigorous application of information garnered from the fossil record to forward-thinking conservation efforts. Their thinking goes like this: If conservationists reach back in history far enough, the past will suggest not only how ecosystems were once composed, but how they could best function in the future. Those at the meeting also said that conservationists must take a wider view of nature than they may have in the past. This still means, in many cases, saving individual species or attempting to maintain some ecosystems much as they are, which is how conservation is generally perceived. But it also means accepting that not all human uses of the environment are inherently bad. "Changed landscapes aren't necessarily trashed landscapes," said Anthony Barnosky, executive director of Stanford's Jasper Ridge Biological Preserve, professor emeritus of integrative biology at UC Berkeley and senior author of the Science paper. "These landscapes can actually be used to help in nature conservation. The question is, how do you best do that?" Also among the Science co-authors is Alexis Mychajliw, a Stanford biology researcher who does fieldwork in the Dominican Republic and whose work exemplifies the new recommended approach to conservation. She explores the fossil record of the island nation to see what its ecosystems were like both before European settlers and before any human habitation. The comparisons between these previous states and the present day can be striking. "Where I work, you only see two living terrestrial mammal species," said Mychajliw, a graduate student in the laboratory of Stanford biology professor Elizabeth Hadly, who is also a co-author of the paper. "But if you dig, actually dig in the dirt, you'll find there were once 25 species of mammals native to this area." This is just one instance of how the fossil record can fill in blanks in natural history that conservationists may be unaware are even missing. Rather than rely on reports or conjecture about what the natural world was like 500 years ago, specialists can piece together what it was like during various periods stretching back millions of years. Given that some natural cycles work on these grand timescales, going back even further than centuries helps explain how cycles may overlap and interact now and in the future. In turn, that can reveal apparent shifts in the natural world that are worthy of concern. Having access to what an ecosystem looked like at different periods in the past also suggests options for creating, achieving and maintaining conservation goals. Already, about 47 percent of ice-free land in the world has been transformed into what ecologists call "novel ecosystems." These ecosystems are unique assemblages of species or systems that didn't exist in pre-industrial times and include cropland, pastureland and timber plantations. The Science paper ("Merging Paleobiology with Conservation Biology to Guide the Future of Terrestrial Ecosystems") points out that these novel ecosystems are unlikely to be restored to what they were before humans. That may not be a bad thing. Instead, the authors suggest that there may be cases where conservationists should embrace novelty to understand how to move forward while supporting both natural diversity and civilization. An example is Stanford's Jasper Ridge Biological Preserve, says Hadly, who is the preserve's faculty director. "Here we have a landscape that humans have used heavily for centuries," she said. "The species that dominate have changed through time, some of them invasives, but it still preserves a remarkable slice of California biodiversity and offers a refuge of nature in the midst of Silicon Valley. Going into the future, we expect more change, but that doesn't lessen its conservation value." Whether an area is novel or historical, the paper argues that conservationists need to carefully consider the services provided by an ecosystem, including air and water purification, carbon sequestration, use of land for agriculture and tourism and the less tangible value of human interactions with the wild. "We rely on nature for almost everything: clean water, food, materials for construction and making computers and phones," said Allison Stegner, postdoctoral research associate at the University of Wisconsin-Madison, former graduate student at UC Berkeley and co-author of the paper. "The pace of global change today is so fast that we stand to lose all of those things that we rely on. Coming up with new approaches to conservation is essential to maintaining human life." The authors say keeping nature diverse, healthy and useful will likely require both historical and novel habitats, efforts that focus on saving species and efforts that don't, input from people who use nature for different purposes and data from very long timescales, some of which can only be obtained through the fossil record. They also said that identifying stakes and stakeholders is a task that goes beyond the sciences. With that in mind, the workshop that led to the paper opened with words from non-scientists, including a senior policy adviser to California Gov. Jerry Brown and a fiction writer who is also the mother of one of the researchers. "They fostered a more humanist approach to our collaboration that lingers," said Hadly. Co-authors of this paper include Patrick Gonzalez, Jason Head, P. David Polly, A. Michelle Lawing, Jussi T. Eronen, David D. Ackerly, Ken Alex, Eric Biber, Jessica Blois, Justin Brashares, Gerardo Ceballos, Edward Davis, Gregory P. Dietl, Rodolfo Dirzo, Holly Doremus, Mikael Fortelius, Harry Greene, Jessica Hellmann, Thomas Hickler, Stephen T. Jackson, Melissa Kemp, Paul L. Koch, Claire Kremen, Emily L. Lindsey, Cindy Looy, Charles R. Marshall, Chase Mendenhall, Andreas Mulch, Carsten Nowak, Uma Ramakrishnan, Jan Schnitzler, Kashish Das Shrestha, Katherine Solari, Lynn Stegner, Nils Chr. Stenseth, Marvalee H. Wake, and Zhibin Zhang. During this research, Barnosky was in the Department of Integrative Biology and Museums of Paleontology and Vertebrate Zoology, University of California, Berkeley. Hadly is also the Paul S. and Billie Achilles Professor in Environmental Biology, senior fellow at the Stanford Woods Institute for the Environment, faculty director of Jasper Ridge Biological Preserve and a member of Stanford Bio-X. Stegner received her B.S. in Ecology and Evolutionary Biology from Stanford in 2010. This workshop was funded by the Integrative Climate Change Biology Group, International Union of Biological Sciences; the Museum of Paleontology, Berkeley Initiative for Global Change Biology, and Office of the Vice Chancellor for Research at the University of California, Berkeley; the Conservation Paleobiology Group at the Department of Biology, Stanford University; and the Senckenberg Biodiversity and Climate Research Centre, Frankfurt, Germany. Additional funding for this work came from the National Science Foundation.


News Article | November 21, 2016
Site: www.eurekalert.org

By charting the slopes and crags on animals' teeth as if they were mountain ranges, scientists at the Smithsonian's National Museum of Natural History have created a powerful new way to learn about the diets of extinct animals from the fossil record. Understanding the diets of animals that lived long ago can tell researchers about the environments they lived in and help them piece together a picture of how the planet has changed over deep time. The new quantitative approach to analyzing dentition, reported Nov. 21 in the journal Methods in Ecology and Evolution, will also give researchers a clearer picture of how animals evolve in response to changes in their environment. "The new method gives researchers a way to measure changes that arose as animals adapted to environments altered by mass extinctions or major climate shifts," said Smithsonian paleontologist Sílvia Pineda-Munoz, who led the technique's development. "By using shape algorithms to examine teeth before and after these perturbations, we can understand the morphological adaptations that happen when there is an [environmental] change. That in turn can help researchers and conservationists predict and plan for such events in the future. It is another tool we can use to understand how present-day communities are going to be affected if something like that happens now." Pineda-Munoz is a postdoctoral fellow in the Natural History Museum's Evolution of Terrestrial Ecosystems Program, which brings together researchers from different disciplines to investigate how terrestrial ecosystems are structured and how they have changed over geologic time. She developed the new method of determining an animal's diet in collaboration with colleagues at Arizona State University, Macquarie University, Monash University and the Museum Victoria in Melbourne. Paleobiologists have long compared the shapes of fossil teeth to those of existing animals to make inferences about what prehistoric species ate millions of years ago. The new method builds on this approach but is more informative and precise, computationally comparing the surfaces of an animal's teeth with those of more than 130 present-day mammals. The technique relies on a three-dimensional scan of a set of teeth, which generates a digital model resembling a topographic map of the Earth's surface. GIS (geographic information system) technology is used to analyze the map, mathematically describing several key features that influence how teeth process food. For example, the program measures how often the slope of tooth surfaces change--an indicator of complexity. Diets made up of foods that require a lot of mechanical processing before they are digested, like tough vegetation, are associated with more complex dentition, Pineda-Munoz explains. While she was a graduate student at Macquarie University, Pineda-Munoz mapped the teeth of 134 contemporary mammals, including representatives from each of eight different dietary categories. "Those categories give detailed information about an animal's primary food source, including plants, meat, fruits, grains, insects, fungus or tree saps, with an additional 'generalist' diet category," Pineda-Munoz said. Pineda-Munoz and her colleagues created a database recording six measurable features of tooth topology for the top and bottom sets of teeth from present-day mammals. Variations in those features reflected differences in the animals' diets. For example, pandas, whose teeth must crush tough leaves, have the most complex teeth, whereas hyenas' scissor-like teeth are efficient for tearing meat. To determine what types of food extinct animals were best equipped to eat, researchers can scan teeth from the fossil record and mathematically compare how their shapes relate to the teeth of animals with known diets--an approach similar to the algorithms websites use to predict what related content a user will enjoy based on past favorites. "Because the method precisely measures the shape of teeth, it will be valuable in assessing how animals' teeth have changed over the course of evolution," Pineda-Munoz said. "It's a method that looks at evolutionary change. It tells you not just what the animal was eating at this point in time, but what the animal was adapted to eating."


Understanding the diets of animals that lived long ago can tell researchers about the environments they lived in and help them piece together a picture of how the planet has changed over deep time. The new quantitative approach to analyzing dentition, reported Nov. 21 in the journal Methods in Ecology and Evolution, will also give researchers a clearer picture of how animals evolve in response to changes in their environment. "The new method gives researchers a way to measure changes that arose as animals adapted to environments altered by mass extinctions or major climate shifts," said Smithsonian paleontologist Sílvia Pineda-Munoz, who led the technique's development. "By using shape algorithms to examine teeth before and after these perturbations, we can understand the morphological adaptations that happen when there is an [environmental] change. That in turn can help researchers and conservationists predict and plan for such events in the future. It is another tool we can use to understand how present-day communities are going to be affected if something like that happens now." Pineda-Munoz is a postdoctoral fellow in the Natural History Museum's Evolution of Terrestrial Ecosystems Program, which brings together researchers from different disciplines to investigate how terrestrial ecosystems are structured and how they have changed over geologic time. She developed the new method of determining an animal's diet in collaboration with colleagues at Arizona State University, Macquarie University, Monash University and the Museum Victoria in Melbourne. Paleobiologists have long compared the shapes of fossil teeth to those of existing animals to make inferences about what prehistoric species ate millions of years ago. The new method builds on this approach but is more informative and precise, computationally comparing the surfaces of an animal's teeth with those of more than 130 present-day mammals. The technique relies on a three-dimensional scan of a set of teeth, which generates a digital model resembling a topographic map of the Earth's surface. GIS (geographic information system) technology is used to analyze the map, mathematically describing several key features that influence how teeth process food. For example, the program measures how often the slope of tooth surfaces change—an indicator of complexity. Diets made up of foods that require a lot of mechanical processing before they are digested, like tough vegetation, are associated with more complex dentition, Pineda-Munoz explains. While she was a graduate student at Macquarie University, Pineda-Munoz mapped the teeth of 134 contemporary mammals, including representatives from each of eight different dietary categories. "Those categories give detailed information about an animal's primary food source, including plants, meat, fruits, grains, insects, fungus or tree saps, with an additional 'generalist' diet category," Pineda-Munoz said. Pineda-Munoz and her colleagues created a database recording six measurable features of tooth topology for the top and bottom sets of teeth from present-day mammals. Variations in those features reflected differences in the animals' diets. For example, pandas, whose teeth must crush tough leaves, have the most complex teeth, whereas hyenas' scissor-like teeth are efficient for tearing meat. To determine what types of food extinct animals were best equipped to eat, researchers can scan teeth from the fossil record and mathematically compare how their shapes relate to the teeth of animals with known diets—an approach similar to the algorithms websites use to predict what related content a user will enjoy based on past favorites. "Because the method precisely measures the shape of teeth, it will be valuable in assessing how animals' teeth have changed over the course of evolution," Pineda-Munoz said. "It's a method that looks at evolutionary change. It tells you not just what the animal was eating at this point in time, but what the animal was adapted to eating." Explore further: Tooth wear patterns suggest Paranthropus early hominins had softer diets than expected More information: Silvia Pineda-Munoz et al, Inferring diet from dental morphology in terrestrial mammals, Methods in Ecology and Evolution (2016). DOI: 10.1111/2041-210X.12691


Clemente C.J.,University of Queensland | Withers P.C.,University of Western Australia | Thompson G.G.,Terrestrial Ecosystems | Lloyd D.,Edith Cowan University | Lloyd D.,Griffith University
Journal of Experimental Biology | Year: 2013

Adaptations promoting greater performance in one habitat are thought to reduce performance in others. However, there are many examples of animals in which, despite habitat differences, such predicted differences in performance do not occur. One such example is the relationship between locomotory performance to habitat for varanid lizards. To explain the lack of difference in locomotor performance we examined detailed observations of the kinematics of each lizard's stride. Differences in kinematics were greatest between climbing and non-climbing species. For terrestrial lizards, the kinematics indicated that increased femur adduction, femur rotation and ankle angle all contributed positively to changes in stride length, but they were constrained for climbing species, probably because of biomechanical restrictions on the centre of mass height (to increase stability on vertical surfaces). Despite climbing species having restricted stride length, no differences have been previously reported in sprint speed between climbing and non-climbing varanids. This is best explained by climbing varanids using an alternative speed modulation strategy of varying stride frequency to avoid the potential trade-off of speed versus stability on vertical surfaces. Thus, by measuring the relevant biomechanics for lizard strides, we have shown how kinematic differences among species can mask performance differences typically associated with habitat variation. © 2013. Published by The Company of Biologists Ltd.


Clemente C.J.,Harvard University | Withers P.C.,University of Western Australia | Thompson G.,Terrestrial Ecosystems | Thompson G.,Edith Cowan University
Physiological and Biochemical Zoology | Year: 2012

Studies of locomotor performance often link variation in morphology with ecology. While maximum sprint speed is a commonly used performance variable, the absolute limits for this performance trait are not completely understood. Absolute maximal speed has often been shown to increase linearly with body size, but several comparative studies covering a large range of body sizes suggest that maximal speed does not increase indefinitely with body mass but rather reaches an optimum after which speed declines. Because of the comparative nature of these studies, it is difficult to determine whether this decrease is due to biomechanical constraints on maximal speed or is a consequence of phylogenetic inertia or perhaps relaxed selection for lower maximal speed at large body size. To explore this issue, we have examined intraspecific variations in morphology, maximal sprint speed, and kinematics for the yellowspotted monitor lizard Varanus panoptes, which varied in body mass from 0.09 to 5.75 kg. We show a curvilinear relationship between body size and absolute maximal sprint speed with an optimal body mass with respect to speed of 1.245 kg. This excludes the phylogenetic inertia hypothesis, because this effect should be absent intraspecifically, while supporting the biomechanical constraints hypothesis. The relaxed selection hypothesis cannot be excluded if there is a size-based behavioral shift intraspecifically, but the biomechanical constraints hypothesis is better supported from kinematic analyses. Kinematic measurements of hind limb movement suggest that the distance moved by the body during the stance phase may limit maximum speed. This limit is thought to be imposed by a decreased ability of the bones and muscles to support body mass for larger lizards. © 2012 by The University of Chicago.


Thompson G.G.,Terrestrial Ecosystems | Thompson G.G.,Edith Cowan University | Thompson S.A.,Terrestrial Ecosystems
Australian Mammalogy | Year: 2014

Mulgaras (Dasycercus cristicauda and D. blythi) are protected by state and commonwealth environmental statutes; as a consequence, land developers and mining companies have an obligation to avoid, mitigate or minimise impacts on these species when they occur in their area of operation (i.e. to implement trapping and translocation programs). Here we assess the effectiveness of searching and trapping programs for mulgaras in four case studies and provide management recommendations to improve outcomes for these species. © 2014 Australian Mammal Society.


Thompson S.,Terrestrial Ecosystems | Thompson G.,Terrestrial Ecosystems | Sackmann J.,Mount Gibson Iron Ltd | Spark J.,59 King William Street | Brown T.,59 King William Street
Pacific Conservation Biology | Year: 2015

The threatened malleefowl (Leipoa ocellata) constructs a large (often > 3 m) incubator mound (nest) that is considered a useful proxy for surveying its presence and abundance in the context of an environmental impact assessment. Here we report on the effectiveness and relative cost of using high-definition aerial photography to search in 3D for malleefowl mounds by comparing results to those of earlier ground-based searches. High-definition colour aerial photography was taken of an area of 7014 ha and searched in 3D for malleefowl mounds. All 24 active (i.e. in use) malleefowl mounds known before the examination of aerial photography were detected using the new assessment technique. Of the 108 total mounds (active and inactive) known from earlier on-ground surveys, 94 (87%) were recorded using the new technique. Mounds not detected were all old and weathered, many barely above ground level and some with vegetation growing in the crater. Approximately 6.3% of the identifications considered 'confident' and 35.0% considered 'potential' based on the aerial photography proved to be false positives. The cost of detecting malleefowl mounds using the interpretation of high-definition 3D colour aerial photography and then subsequently examining these areas on the ground is appreciably cheaper than on-ground grid searches. © CSIRO 2015.


Thompson S.A.,Terrestrial Ecosystems | Thompson G.G.,Terrestrial Ecosystems
Pacific Conservation Biology | Year: 2015

Vegetation clearing is often a precursor to urban, industrial and mining developments. Vertebrate fauna are often lost and injured during this process; however, these impacts are often mitigated by implementing a fauna rescue program. Here we report on the success of a trapping and relocation program and the use of fauna rescue personnel to remove vertebrate fauna from two vegetation-clearing programs. We provide comment on the impact of various machines that are used in the clearing process and which taxa have the higher survival rates, and conclude with some management recommendations that will provide better outcomes for vertebrate fauna during vegetation-clearing programs. © CSIRO 2015.


Thompson G.G.,Terrestrial Ecosystems | Thompson S.A.,Terrestrial Ecosystems
Pacific Conservation Biology | Year: 2015

In all, 154 of 158 above-ground termitaria deconstructed in the Pilbara of Western Australia contained at least one vertebrate, and there was a mean of 30.4 (s.e.= 2.03) vertebrates and 4.5 (s.e.= 0.17) species in each mound. There was a significant difference in the relative abundance of species found in the termitaria and the 64 species found in the adjacent area. Termitaria were mostly occupied by eight species: Gehyra pilbara (66.3% of captures), Heteronotia binoei (13.7% of captures), Furina ornata (6.9%), Antaresia stimsoni (3.3%), Cyclorana maini (3.0%), Gehyra variegata (1.5%), Suta punctata (1.3%) and Planigale sp. (0.9%). It is likely that F. ornata, A. stimsoni and S. punctata used termitaria as a diurnal refuge and also prey upon reptiles living in the mound. If other termitaria in the Pilbara support a similarly high number of vertebrates, then these mounds provide an environmentally significant microhabitat and vertebrate fauna inhabiting the mounds should be captured and relocated before the termitaria are cleared or isolated as a result of development. © CSIRO 2015.

Loading Terrestrial Ecosystems collaborators
Loading Terrestrial Ecosystems collaborators