Applying collocation and APRIORI analyses to chimpanzee diets: Methods for investigating nonrandom food combinations in primate self-medication.

Identifying novel medicinal resources in chimpanzee diets has historically presented challenges, requiring extensive behavioral data collection and health monitoring, accompanied by expensive pharmacological analyses. When putative therapeutic self-medicative behaviors are observed, these events are often considered isolated occurrences, with little attention paid to other resources ingested in combination. For chimpanzees, medicinal resource combinations could play an important role in maintaining well-being by tackling different symptoms of an illness, chemically strengthening efficacy of a treatment, or providing prophylactic compounds that prevent future ailments. We call this concept the self-medicative resource combination hypothesis. However, a dearth of methodological approaches for holistically investigating primate feeding ecology has limited our ability to identify nonrandom resource combinations and explore potential synergistic relationships between medicinal resource candidates. Here we present two analytical tools that test such a hypothesis and demonstrate these approaches on feeding data from the Sonso chimpanzee community in Budongo Forest, Uganda. Using 4 months of data, we establish that both collocation and APRIORI analyses are effective exploratory tools for identifying binary combinations, and that APRIORI is effective for multi-item rule associations. We then compare outputs from both methods, finding up to 60% agreement, and propose APRIORI as more effective for studies requiring control over confidence intervals and those investigating nonrandom associations between more than two resources. These analytical tools, which can be extrapolated across the animal kingdom, can provide a cost-effective and efficient method for targeting resources for further pharmacological investigation, potentially aiding in the discovery of novel medicines.

The first Miocene fossils from coastal woodlands in the southern East African Rift.

The Miocene was a key time in the evolution of African ecosystems witnessing the origin of the African apes and the isolation of eastern coastal forests through an expanding arid corridor. Until recently, however, Miocene sites from the southeastern regions of the continent were unknown. Here, we report the first Miocene fossil teeth from the shoulders of the Urema Rift in Gorongosa National Park, Mozambique. We provide the first 1) radiometric ages of the Mazamba Formation, 2) reconstructions of paleovegetation in the region based on pedogenic carbonates and fossil wood, and 3) descriptions of fossil teeth. Gorongosa is unique in the East African Rift in combining marine invertebrates, marine vertebrates, reptiles, terrestrial mammals, and fossil woods in coastal paleoenvironments. The Gorongosa fossil sites offer the first evidence of woodlands and forests on the coastal margins of southeastern Africa during the Miocene, and an exceptional assemblage of fossils including new species.

Unsupervised learning of satellite images enhances discovery of late Miocene fossil sites in the Urema Rift, Gorongosa, Mozambique.

BACKGROUND: Paleoanthropological research focus still devotes most resources to areas generally known to be fossil rich instead of a strategy that first maps and identifies possible fossil sites in a given region. This leads to the paradoxical task of planning paleontological campaigns without knowing the true extent and likely potential of each fossil site and, hence, how to optimize the investment of time and resources. Yet to answer key questions in hominin evolution, paleoanthropologists must engage in fieldwork that targets substantial temporal and geographical gaps in the fossil record. How can the risk of potentially unsuccessful surveys be minimized, while maximizing the potential for successful surveys? METHODS: Here we present a simple and effective solution for finding fossil sites based on clustering by unsupervised learning of satellite images with the k-means algorithm and pioneer its testing in the Urema Rift, the southern termination of the East African Rift System (EARS). We focus on a relatively unknown time period critical for understanding African apes and early hominin evolution, the early part of the late Miocene, in an overlooked area of southeastern Africa, in Gorongosa National Park, Mozambique. This clustering approach highlighted priority targets for prospecting that represented only 4.49% of the total area analysed. RESULTS: Applying this method, four new fossil sites were discovered in the area, and results show an 85% accuracy in a binary classification. This indicates the high potential of a remote sensing tool for exploratory paleontological surveys by enhancing the discovery of productive fossiliferous deposits. The relative importance of spectral bands for clustering was also determined using the random forest algorithm, and near-infrared was the most important variable for fossil site detection, followed by other infrared variables. Bands in the visible spectrum performed the worst and are not likely indicators of fossil sites. DISCUSSION: We show that unsupervised learning is a useful tool for locating new fossil sites in relatively unexplored regions. Additionally, it can be used to target specific gaps in the fossil record and to increase the sample of fossil sites. In Gorongosa, the discovery of the first estuarine coastal forests of the EARS fills an important paleobiogeographic gap of Africa. These new sites will be key for testing hypotheses of primate evolution in such environmental settings.

CADOES: An interactive machine-learning approach for sex estimation with the pelvis.

The pelvis is consistently regarded as the most sexually dimorphic region of the human skeleton, and methods for sex estimation with the pelvic bones are usually very accurate. In this investigation, population-specific osteometric models for the assessment of sex with the pelvis were designed using a dataset provided by J.A. Serra (1938) that included 256 individuals (131 females and 125 males) from the Coimbra Identified Skeletal Collection and 38 metric variables. The models for sex estimation were operationalized through an online application and decision support system, CADOES. Different classification algorithms generated high accuracy models, ranging from 85% to 92%, with only three variables; and from 85.33% to 97.33%, with all 38 variables. CADOES conveys a probabilistic prediction of skeletal sex, as well as a suite of attributes with educational applicability in the fields of human skeletal anatomy and statistics. This study upholds the value of the pelvis for the estimation of skeletal sex and provides models for that can be applied with high accuracy and low bias.

Outline Shape Analysis on the Trochlear Constriction and Olecranon Fossa of the Humerus: Insights for Sex Estimation and a New Computational Tool.

Sex estimation through visual analysis of the distal humerus can contribute to establishing the biological profile of an unidentified skeletal individual. Using statistical shape modeling, the trochlear constriction open curves and olecranon fossa closed outlines of 151 humeri were digitized and analyzed. The shape configurations exhibited strikingly different degrees of sexual dimorphism when evaluated using linear discriminant analysis with leave-one-out cross-validation. The trochlear constriction performed poorly, correctly classifying 63.6% of the individuals. However, the olecranon fossa showed high sexual dimorphism, presenting a 94.0% accuracy. A simpler model using only two principal components was also generated. While the accuracy is slightly inferior (88.1%), it has the advantage of being constrained to bidimensional components that were translated into morphoscopic variables within a simulator interface. This allowed us to implement the method through a web application that does not require users to be trained in landmark digitization or have knowledge of geometric morphometrics.

One for all and all for one: Linear regression from the mass of individual bones to assess human skeletal mass completeness.

OBJECTIVES: Complete and accurate human skeletal inventory is seldom possible in archaeological and forensic cases involving severe fragmentation. In such cases, skeletal mass comparisons with published references may be used as an alternative to assess skeletal completeness but they are too general for a case-by-case routine analysis. The objective is to solve this issue by creating linear regression equations to estimate the total mass of a skeleton based on the mass of individual bones. MATERIALS AND METHODS: Total adult skeletal mass and individual mass of the clavicle, humerus, femur, patella, carpal, metacarpal, tarsal, and metatarsal bones were recorded in a sample of 60 skeletons from the 21st century identified skeletal collection (University of Coimbra). The sample included 32 females and 28 males with ages ranging from 31 to 96 years (mean = 76.4; sd = 14.8). Skeletal mass linear regression equations were calculated based on this sample. RESULTS: The mass of individual bones was successfully used to predict the approximate total mass of the adult skeleton. The femur, humerus, and second metacarpal were the best predictors of total skeletal mass with root mean squared errors ranging from 292.9 to 346.1 g. DISCUSSION: Linear regression was relatively successful at estimating adult skeletal mass. The non-normal distribution of the sample in terms of mass may have reduced the predictive power of the equations. These results have clear impact for bioanthropology, especially forensic anthropology, since this method may provide better estimates of the completeness of the skeleton or the minimum number of individuals. Am J Phys Anthropol 160:427-432, 2016. © 2016 Wiley Periodicals, Inc.