Functional traits of fossil plants.

A minuscule fraction of the Earth's paleobiological diversity is preserved in the geological record as fossils. What plant remnants have withstood taphonomic filtering, fragmentation, and alteration in their journey to become part of the fossil record provide unique information on how plants functioned in paleo-ecosystems through their traits. Plant traits are measurable morphological, anatomical, physiological, biochemical, or phenological characteristics that potentially affect their environment and fitness. Here, we review the rich literature of paleobotany, through the lens of contemporary trait-based ecology, to evaluate which well-established extant plant traits hold the greatest promise for application to fossils. In particular, we focus on fossil plant functional traits, those measurable properties of leaf, stem, reproductive, or whole plant fossils that offer insights into the functioning of the plant when alive. The limitations of a trait-based approach in paleobotany are considerable. However, in our critical assessment of over 30 extant traits we present an initial, semi-quantitative ranking of 26 paleo-functional traits based on taphonomic and methodological criteria on the potential of those traits to impact Earth system processes, and for that impact to be quantifiable. We demonstrate how valuable inferences on paleo-ecosystem processes (pollination biology, herbivory), past nutrient cycles, paleobiogeography, paleo-demography (life history), and Earth system history can be derived through the application of paleo-functional traits to fossil plants.

Semantic segmentation of plant roots from RGB (mini-) rhizotron images-generalisation potential and false positives of established methods and advanced deep-learning models.

BACKGROUND: Manual analysis of (mini-)rhizotron (MR) images is tedious. Several methods have been proposed for semantic root segmentation based on homogeneous, single-source MR datasets. Recent advances in deep learning (DL) have enabled automated feature extraction, but comparisons of segmentation accuracy, false positives and transferability are virtually lacking. Here we compare six state-of-the-art methods and propose two improved DL models for semantic root segmentation using a large MR dataset with and without augmented data. We determine the performance of the methods on a homogeneous maize dataset, and a mixed dataset of > 8 species (mixtures), 6 soil types and 4 imaging systems. The generalisation potential of the derived DL models is determined on a distinct, unseen dataset. RESULTS: The best performance was achieved by the U-Net models; the more complex the encoder the better the accuracy and generalisation of the model. The heterogeneous mixed MR dataset was a particularly challenging for the non-U-Net techniques. Data augmentation enhanced model performance. We demonstrated the improved performance of deep meta-architectures and feature extractors, and a reduction in the number of false positives. CONCLUSIONS: Although correction factors are still required to match human labelled root lengths, neural network architectures greatly reduce the time required to compute the root length. The more complex architectures illustrate how future improvements in root segmentation within MR images can be achieved, particularly reaching higher segmentation accuracies and model generalisation when analysing real-world datasets with artefacts-limiting the need for model retraining.

Leaf-level coordination principles propagate to the ecosystem scale.

Fundamental axes of variation in plant traits result from trade-offs between costs and benefits of resource-use strategies at the leaf scale. However, it is unclear whether similar trade-offs propagate to the ecosystem level. Here, we test whether trait correlation patterns predicted by three well-known leaf- and plant-level coordination theories - the leaf economics spectrum, the global spectrum of plant form and function, and the least-cost hypothesis - are also observed between community mean traits and ecosystem processes. We combined ecosystem functional properties from FLUXNET sites, vegetation properties, and community mean plant traits into three corresponding principal component analyses. We find that the leaf economics spectrum (90 sites), the global spectrum of plant form and function (89 sites), and the least-cost hypothesis (82 sites) all propagate at the ecosystem level. However, we also find evidence of additional scale-emergent properties. Evaluating the coordination of ecosystem functional properties may aid the development of more realistic global dynamic vegetation models with critical empirical data, reducing the uncertainty of climate change projections.

High frequency root dynamics: sampling and interpretation using replicated robotic minirhizotrons.

Automating dynamic fine root data collection in the field is a longstanding challenge with multiple applications for co-interpretation and synthesis for ecosystem understanding. High frequency root data are only achievable with paired automated sampling and processing. However, automatic minirhizotron (root camera) instruments are still rare and data are often not collected in natural soils or analysed at high temporal resolution. Instruments must also be affordable for replication and robust under variable natural conditions. Here, we show a system built with off-the-shelf parts which samples at sub-daily resolution. We paired this with a neural network to analyse all images collected. We performed two mesocosm studies and two field trials alongside ancillary data collection (soil CO2 efflux, temperature, and moisture content, and 'PhenoCam'-derived above-ground dynamics). We produce robust and replicated daily time series of root dynamics under all conditions. Temporal root changes were a stronger driver than absolute biomass on soil CO2 efflux in the mesocosm. Proximal sensed above-ground dynamics and below-ground dynamics from minirhizotron data were not synchronized. Root properties extracted were sensitive to soil moisture and occasionally to time of day (potentially relating to soil moisture). This may only affect high frequency imagery and should be considered in interpreting such data.

"It's a waste of time coming here, better go to private clinics with wider options for treatment": patient's perception on dental services provided in Fiji.

BACKGROUND: Oral health can influence the quality of an individual's life. Patient's perception of the service plays a vital role in understanding the reasons as to why a patient may be satisfied or dissatisfied with the service that they accessed at a dental setting. As no studies have been done in Fiji until now, this study aimed to explore the perceptions of patients on services provided by the largest dental clinic in the Central/Eastern division in Fiji. METHOD: A qualitative study was used to collect data from patients who visited the Colonial War Memorial Hospital (CWMH) dental clinic in Suva Fiji from 5th August to 31st October, 2020. All patients above the age of 18 of both genders and from any ethnicity who visited the CWMH dental clinic during the period of data collection were included the study. A total of 25 participants were interviewed for this study using the in-depth interview method till data saturation occurred. A semi-structured open-ended questionnaire was used to collect data using face-to-face in-depth interviews. The data were transcribed and analyzed using manual thematic analysis process to gather the themes and sub-themes for the results. RESULTS: A total of 25 patients were interviewed, with a majority (n = 14) being men and 15 were of I-Taukei background. Five themes emerging from data analysis include: Waiting time before treatment, Cost of Treatment, Accessibility of services, Privacy and confidentiality and Range of treatment options. The patients had an expectation to get the best treatment but face many hurdles while trying to get the treatment that they expect. The shortfalls on the part of this dental clinic caused an onset of negative perception among its users. CONCLUSION: This study shows an overall dissatisfaction with regards to services delivery among the patients who use the CWMH dental clinic for dental care purposes. The decision makers need to look into the genuine concerns that have been raised by patients in order to create improvements in services delivery and create an array of satisfaction for its patients.

Long-term ecosystem nitrogen limitation from foliar δ15 N data and a land surface model.

The effect of nutrient availability on plant growth and the terrestrial carbon sink under climate change and elevated CO2 remains one of the main uncertainties of the terrestrial carbon cycle. This is partially due to the difficulty of assessing nutrient limitation at large scales over long periods of time. Consistent declines in leaf nitrogen (N) content and leaf δ15 N have been used to suggest that nitrogen limitation has increased in recent decades, most likely due to the concurrent increase in atmospheric CO2 . However, such data sets are often not straightforward to interpret due to the complex factors that contribute to the spatial and temporal variation in leaf N and isotope concentration. We use the land surface model (LSM) QUINCY, which has the unique capacity to represent N isotopic processes, in conjunction with two large data sets of foliar N and N isotope content. We run the model with different scenarios to test whether foliar δ15 N isotopic data can be used to infer large-scale N limitation and if the observed trends are caused by increasing atmospheric CO2 , changes in climate or changes in sources and magnitude of anthropogenic N deposition. We show that while the model can capture the observed change in leaf N content and predict widespread increases in N limitation, it does not capture the pronounced, but very spatially heterogeneous, decrease in foliar δ15 N observed in the data across the globe. The addition of an observation-based temporal trend in isotopic composition of N deposition leads to a more pronounced decrease in simulated leaf δ15 N. Our results show that leaf δ15 N observations cannot, on their own, be used to assess global-scale N limitation and that using such a data set in conjunction with an LSM can reveal the drivers behind the observed patterns.

Global sensitivities of forest carbon changes to environmental conditions.

The responses of forest carbon dynamics to fluctuations in environmental conditions at a global scale remain elusive. Despite the understanding that favourable environmental conditions promote forest growth, these responses have been challenging to observe across different ecosystems and climate gradients. Based on a global annual time series of aboveground biomass (AGB) estimated from radar satellites between 1992 and 2018, we present forest carbon changes and provide insights on their sensitivities to environmental conditions across scales. Our findings indicate differences in forest carbon changes across AGB classes, with regions with carbon stocks of 50-125 MgC ha-1 depict the highest forest carbon gains and losses, while regions with 125-150 MgC ha-1  have the lowest forest carbon gains and losses in absolute terms. Net forest carbon change estimates show that the arc-of-deforestation and the Congo Basin were the main hotspots of forest carbon loss, while a substantial part of European forest gained carbon during the last three decades. Furthermore, we observe that changes in forest carbon stocks were systematically positively correlated with changes in forest cover fraction. At the same time, it was not necessarily the case with other environmental variables, such as air temperature and water availability at the bivariate level. We also used a model attribution method to demonstrate that atmospheric conditions were the dominant control of forest carbon changes (56% of the total study area) followed by water-related (29% of the total study area) and vegetation (15% of the total study area) conditions. Regionally, we find evidence that carbon gains from long-term forest growth covary with long-term carbon sinks inferred from atmospheric inversions. Our results describe the contributions from the atmosphere, water-related and vegetation conditions to forest carbon changes and provide new insights into the underlying mechanisms of the coupling between forest growth and the global carbon cycle.

Influence of medical domain knowledge on deep learning for Alzheimer's disease prediction.

BACKGROUND AND OBJECTIVE: Alzheimer's disease (AD) is the most common type of dementia that can seriously affect a person's ability to perform daily activities. Estimates indicate that AD may rank third as a cause of death for older people, after heart disease and cancer. Identification of individuals at risk for developing AD is imperative for testing therapeutic interventions. The objective of the study was to determine could diagnostics of AD from EMR data alone (without relying on diagnostic imaging) be significantly improved by applying clinical domain knowledge in data preprocessing and positive dataset selection rather than setting naïve filters. METHODS: Data were extracted from the repository of heterogeneous ambulatory EMR data, collected from primary care medical offices all over the U.S. Medical domain knowledge was applied to build a positive dataset from data relevant to AD. Selected Clinically Relevant Positive (SCRP) datasets were used as inputs to a Long-Short-Term Memory (LSTM) Recurrent Neural Network (RNN) deep learning model to predict will the patient develop AD. RESULTS: Risk scores prediction of AD using the drugs domain information in an SCRP AD dataset of 2,324 patients achieved high out-of-sample score - 0.98-0.99 Area Under the Precision-Recall Curve (AUPRC) when using 90% of SCRP dataset for training. AUPRC dropped to 0.89 when training the model using less than 1,500 cases from the SCRP dataset. The model was still significantly better than when using naïve dataset selection. CONCLUSION: The LSTM RNN method that used data relevant to AD performed significantly better when learning from the SCRP dataset than when datasets were selected naïvely. The integration of qualitative medical knowledge for dataset selection and deep learning technology provided a mechanism for significant improvement of AD prediction. Accurate and early prediction of AD is significant in the identification of patients for clinical trials, which can possibly result in the discovery of new drugs for treatments of AD. Also, the contribution of the proposed predictions of AD is a better selection of patients who need imaging diagnostics for differential diagnosis of AD from other degenerative brain disorders.

Nutrients and water availability constrain the seasonality of vegetation activity in a Mediterranean ecosystem.

Anthropogenic nitrogen (N) deposition and resulting differences in ecosystem N and phosphorus (P) ratios are expected to impact photosynthetic capacity, that is, maximum gross primary productivity (GPPmax ). However, the interplay between N and P availability with other critical resources on seasonal dynamics of ecosystem productivity remains largely unknown. In a Mediterranean tree-grass ecosystem, we established three landscape-level (24 ha) nutrient addition treatments: N addition (NT), N and P addition (NPT), and a control site (CT). We analyzed the response of ecosystem to altered nutrient stoichiometry using eddy covariance fluxes measurements, satellite observations, and digital repeat photography. A set of metrics, including phenological transition dates (PTDs; timing of green-up and dry-down), slopes during green-up and dry-down period, and seasonal amplitude, were extracted from time series of GPPmax and used to represent the seasonality of vegetation activity. The seasonal amplitude of GPPmax was higher for NT and NPT than CT, which was attributed to changes in structure and physiology induced by fertilization. PTDs were mainly driven by rainfall and exhibited no significant differences among treatments during the green-up period. Yet, both fertilized sites senesced earlier during the dry-down period (17-19 days), which was more pronounced in the NT due to larger evapotranspiration and water usage. Fertilization also resulted in a faster increase in GPPmax during the green-up period and a sharper decline in GPPmax during the dry-down period, with less prominent decline response in NPT. Overall, we demonstrated seasonality of vegetation activity was altered after fertilization and the importance of nutrient-water interaction in such water-limited ecosystems. With the projected warming-drying trend, the positive effects of N fertilization induced by N deposition on GPPmax may be counteracted by an earlier and faster dry-down in particular in areas where the N:P ratio increases, with potential impact on the carbon cycle of water-limited ecosystems.

Decomposition nitrogen is better retained than simulated deposition from mineral amendments in a temperate forest.

Nitrogen (N) deposition (NDEP ) drives forest carbon (C) sequestration but the size of this effect is still uncertain. In the field, an estimate of these effects can be obtained by applying mineral N fertilizers over the soil or forest canopy. A 15 N label in the fertilizer can be then used to trace the movement of the added N into ecosystem pools and deduce a C effect. However, N recycling via litter decomposition provides most of the nutrition for trees, even under heavy NDEP inputs. If this recycled litter nitrogen is retained in ecosystem pools differently to added mineral N, then estimates of the effects of NDEP on the relative change in C (∆C/∆N) based on short-term isotope-labelled mineral fertilizer additions should be questioned. We used 15 N labelled litter to track decomposed N in the soil system (litter, soils, microbes, and roots) over 18 months in a Sitka spruce plantation and directly compared the fate of this 15 N to an equivalent amount in simulated NDEP treatments. By the end of the experiment, three times as much 15 N was retained in the O and A soil layers when N was derived from litter decomposition than from mineral N additions (60% and 20%, respectively), primarily because of increased recovery in the O layer. Roots expressed slightly more 15 N tracer from litter decomposition than from simulated mineral NDEP (7.5% and 4.5%) and compared to soil recovery, expressed proportionally more 15 N in the A layer than the O layer, potentially indicating uptake of organic N from decomposition. These results suggest effects of NDEP on forest ∆C/∆N may not be apparent from mineral 15 N tracer experiments alone. Given the importance of N recycling, an important but underestimated effect of NDEP is its influence on the rate of N release from litter.

Does canopy nitrogen uptake enhance carbon sequestration by trees?

Temperate forest (15) N isotope trace experiments find nitrogen (N) addition-driven carbon (C) uptake is modest as little additional N is acquired by trees; however, several correlations of ambient N deposition against forest productivity imply a greater effect of atmospheric nitrogen deposition than these studies. We asked whether N deposition experiments adequately represent all processes found in ambient conditions. In particular, experiments typically apply (15) N to directly to forest floors, assuming uptake of nitrogen intercepted by canopies (CNU) is minimal. Additionally, conventional (15) N additions typically trace mineral (15) N additions rather than litter N recycling and may increase total N inputs above ambient levels. To test the importance of CNU and recycled N to tree nutrition, we conducted a mesocosm experiment, applying 54 g N/(15) N ha(-1)  yr(-1) to Sitka spruce saplings. We compared tree and soil (15) N recovery among treatments where enrichment was due to either (1) a (15) N-enriched litter layer, or mineral (15) N additions to (2) the soil or (3) the canopy. We found that 60% of (15) N applied to the canopy was recovered above ground (in needles, stem and branches) while only 21% of (15) N applied to the soil was found in these pools. (15) N recovery from litter was low and highly variable. (15) N partitioning among biomass pools and age classes also differed among treatments, with twice as much (15) N found in woody biomass when deposited on the canopy than soil. Stoichiometrically calculated N effect on C uptake from (15) N applied to the soil, scaled to real-world conditions, was 43 kg C kg N(-1) , similar to manipulation studies. The effect from the canopy treatment was 114 kg C kg N(-1) . Canopy treatments may be critical to accurately represent N deposition in the field and may address the discrepancy between manipulative and correlative studies.

Stem injection of 15N-NH4NO3 into mature Sitka spruce (Picea sitchensis).

Stem injection techniques can be used to introduce (15)N into trees to overcome a low variation in natural abundance and label biomass with a distinct (15)N signature, but have tended to target small and young trees, of a variety of species, with little replication. We injected 98 atom% (15)N ammonium nitrate (NH4NO3) solution into 13 mature, 9- to 13-m tall edge-profile Sitka spruce trees in order to produce a large quantity of labelled litter, examining the distribution of the isotope throughout the canopy after felling in terms of both total abundance of (15)N and relative distribution of the isotope throughout individual trees. Using a simple mass balance of the canopy alone, based on observed total needle biomass and modelled branch biomass, all of the isotope injected was accounted for, evenly split between needles and branches, but with a high degree of variability both within individual trees, and among trees. Both (15)N abundance and relative within-canopy distribution were biased towards the upper and middle crown in foliage. Recovery of the label in branches was much more variable than in needles, possibly due to differences in nitrogen allocation for both growth and storage, which differ seasonally between foliage and woody biomass.