Among grapevine diseases affecting European vineyards, Flavescence dorée (FD) and Grapevine Trunk Diseases (GTD) are considered the most relevant challenges for viticulture because of the damage they ...cause to vineyards. Unmanned Aerial Vehicle (UAV) multispectral imagery could be a powerful tool for the automatic detection of symptomatic vines. However, one major difficulty is to discriminate different kinds of diseases leading to similar leaves discoloration as it is the case with FD and GTD for red vine cultivars. The objective of this paper is to evaluate the potentiality of UAV multispectral imagery to separate: symptomatic vines including FD and GTD (Esca and black dead arm) from asymptomatic vines (Case 1) and FD vines from GTD ones (Case 2). The study sites are localized in the Gaillac and Minervois wine production regions (south of France). A set of seven vineyards covering five different red cultivars was studied. Field work was carried out between August and September 2016. In total, 218 asymptomatic vines, 502 FD vines and 199 GTD vines were located with a centimetric precision GPS. UAV multispectral images were acquired with a MicaSense RedEdge® sensor and were processed to ultimately obtain surface reflectance mosaics at 0.10 m ground spatial resolution. In this study, the potentiality of 24 variables (5 spectral bands, 15 vegetation indices and 4 biophysical parameters) are tested. The vegetation indices are selected for their potentiality to detect abnormal vegetation behavior in relation to stress or diseases. Among the biophysical parameters selected, three are directly linked to the leaf pigments content (chlorophyll, carotenoid and anthocyanin). The first step consisted in evaluating the performance of the 24 variables to separate symptomatic vine vegetation (FD or/and GTD) from asymptomatic vine vegetation using the performance indicators from the Receiver Operator Characteristic (ROC) Curve method (i.e., Area Under Curve or AUC, sensibility and specificity). The second step consisted in mapping the symptomatic vines (FD and/or GTD) at the scale of the field using the optimal threshold resulting from the ROC curve. Ultimately, the error between the level of infection predicted by the selected variables (proportion of symptomatic pixels by vine) and observed in the field (proportion of symptomatic leaves by vine) is calculated. The same methodology is applied to the three levels of analysis: by vineyard, by cultivar (Gamay, Fer Servadou) and by berry color (all red cultivars). At the vineyard and cultivar levels, the best variables selected varies. The AUC of the best vegetation indices and biophysical parameters varies from 0.84 to 0.95 for Case 1 and 0.74 to 0.90 for Case 2. At the berry color level, no variable is efficient in discriminating FD vines from GTD ones (Case 2). For Case 1, the best vegetation indices and biophysical parameter are Red Green Index (RGI)/ Green-Red Vegetation Index (GRVI) (based on the green and red spectral bands) and Car (linked to carotenoid content). These variables are more effective in mapping vines with a level of infection greater than 50%. However, at the scale of the field, we observe misclassified pixels linked to the presence of mixed pixels (shade, bare soil, inter-row vegetation and vine vegetation) and other factors of abnormal coloration (e.g., apoplectic vines).
About predictions in spatial autoregressive models: optimal and almost optimal strategies. Spatial Economic Analysis. This paper addresses the problem of prediction in the spatial autoregressive ...(SAR) model for areal data, which is classically used in spatial econometrics. With kriging theory, prediction using the best linear unbiased predictors (BLUPs) is at the heart of the geostatistical literature. From a methodological point of view, we explore the limits of the extension of BLUP formulas in the context of SAR models for out-of-sample prediction simultaneously at several sites. We propose a more tractable 'almost best' alternative and clarify the relationship between the BLUP and a proper expectation-maximization (EM) algorithm predictor. From an empirical perspective, we present data-based simulations to compare the efficiency of classical formulas with the best and almost best predictions.
Flavescence dorée is a grapevine disease affecting European vineyards which has severe economic consequences and containing its spread is therefore considered as a major challenge for viticulture. ...Flavescence dorée is subject to mandatory pest control including removal of the infected vines and, in this context, automatic detection of Flavescence dorée symptomatic vines by unmanned aerial vehicle (UAV) remote sensing could constitute a key diagnosis instrument for growers. The objective of this paper is to evaluate the feasibility of discriminating the Flavescence dorée symptoms in red and white cultivars from healthy vine vegetation using UAV multispectral imagery. Exhaustive ground truth data and UAV multispectral imagery (visible and near-infrared domain) have been acquired in September 2015 over four selected vineyards in Southwest France. Spectral signatures of healthy and symptomatic plants were studied with a set of 20 variables computed from the UAV images (spectral bands, vegetation indices and biophysical parameters) using univariate and multivariate classification approaches. Best results were achieved with red cultivars (both using univariate and multivariate approaches). For white cultivars, results were not satisfactory either for the univariate or the multivariate. Nevertheless, external accuracy assessment show that despite problems of Flavescence dorée and healthy pixel misclassification, an operational Flavescence dorée mapping technique using UAV-based imagery can still be proposed.
▶ Spatial scale mismatch (SSM) affects the efficiency of agri-environmental policies. ▶ A system approach to the landscape is required to resolve spatial scale mismatch. ▶ Solutions to SSM are to be ...found in hierarchy, organization and co-management theories. ▶ Terminology and theoretical frameworks have to be used rigorously.
The difficulty to spatially link the process levels of organizing agricultural management with those of investigating biodiversity preservation creates a spatial scale mismatch which affects the effectiveness of agri-environmental policies. Starting from a literature review this study offers a panorama of the ways authors approach spatial scale mismatch and the solutions they propose to resolve it. We made the hypothesis that the authors rely, sometimes implicitly, on theoretical frameworks to propose their solutions.
Only 15% of the references in which the authors examine the question of spatial scale mismatch show a systemic approach to the question, taking into account simultaneously ecological and managerial processes. We identify two major types of theory linked to the solutions proposed by the authors: those that refer to “multi-scale/multi-level” management for which hierarchy theory and landscape ecology are referred to explicitly; those that imply collective management and coordination, which refer to the theory of organization of biological systems and to social–ecological systems. These theories and their properties imply a change of paradigm which could allow for a better articulation between biodiversity and agricultural management.
Based on this literature search we suggest that the problems in resolving spatial scale mismatch could be due to the fact that: (1) authors generally do not have a systemic approach since they consider ecological and managerial processes separately, and (2) terminology and theoretical frameworks are used inaccurately.
While there are socio-economic difficulties in the implementation of biodiversity conservation programs in agricultural zones, there are also shortcomings linked to the theoretical representation framework. These shortcomings may hinder the articulation between ecological and managerial processes, this is why approaches are suggested here allowing for a better match between the representations of ecological and managerial processes.
•Plot context primarily explains tree-related microhabitats occurrence.•Plot context is a black box combining environmental, management and biotic factors.•From the literature, we identified 21 ...factors that may play a key role in TreM formation.•A sub-set of 9 factors that should be prioritised in the future is suggested.
Tree-related microhabitats (TreMs) have been identified as key features for forest-dwelling taxa and are often employed as measures for biodiversity conservation in integrative forest management. However, managing forests to ensure an uninterrupted resource supply for TreM-dwelling taxa is challenging since TreMs are structures with a limited availability, some of which are triggered by stochastic events or require a long time to develop. At the tree scale, the role of tree species, diameter at breast height (dbh) and status (i.e. living vs standing dead) for favouring TreM occurrence has been quantified and modelled in several studies, since these tree features are routinely recorded in the field. However, TreM occurrence remains difficult to predict, hampering the elaboration of applicable management strategies that consider TreMs. Using an international database encompassing 110,000 trees, we quantified the explanatory power of tree species, dbh, status, time since last harvest and plot context for predicting TreM occurrence at the tree level. Plot context is so far a “black box” that combines local environmental conditions, past and current management legacies, with local biotic features that have high explanatory power for predicting TreM occurrence. Then, based on the literature, we established a set of 21 factors related to site, stand and tree features for which there is a strong assumption that they play a key role in TreM formation. Finally, we identified a sub-set of nine features that should be recorded in the future to provide additional information to enable better prediction of the occurrence of particular TreMs: (i) at plot level: slope, exposure, altitude and presence of cliffs; and (ii) at tree level: bark features, phyllotaxis and compartmentalization capacity of the tree species, plus ontogenic stage and physiological state of the individual tree sampled.
•33 significant “non-random” co-occurrences were highlighted for broadleaves.•9 significant “non-random” co-occurrences were highlighted for conifers.•Mutually exclusive co-occurrences were found for ...conifers only.•Short lists of TreMs are provided to assess TreMs in routine forest management.
A Tree-related Microhabitat (TreM) is a distinct, well-delineated morphological singularity occurring on living or standing dead trees, which constitutes a crucial substrate or life site for various species. TreMs are widely recognized as key features for biodiversity. Current TreM typology identifies 47 TreM types according to their morphology and their associated taxa. In order to provide a range of resolutions and make the typology more user-friendly, these 47 TreM types have been pooled into 15 groups and seven forms. Depending on the accuracy required and the time available, a user can now choose to describe TreMs at resolution levels corresponding to type, group or form. Another way to more easily record TreMs during routine management work would be to use co-occurrence patterns to reduce the number of observed TreMs required. Based on a large international TreM database (2052 plots; 70,958 individual trees; 78 tree species), we evaluated both the significance and the magnitude of TreM co-occurrence on living trees for 11 TreM groups. We highlighted 33 significant co-occurrences for broadleaves and nine for conifers. Bark loss, rot hole, crack and polypore had the highest number of positive co-occurrences (N = 8) with other TreMs on broadleaves; bark loss (N = 4) had the highest number for conifers. We found mutually exclusive occurrences only for conifers: Exposed Heartwood excluded both dendrotelm and sap run. Among the four variables we tested for their positive contribution to significant co-occurrences, tree diameter at breast height was the most consistent. Based on our results and practical considerations, we selected three TreM groups for broadleaves, and nine for conifers, and formed useful short lists to reduce the number of TreM groups to assess during routine forest management work in the field. In addition, detecting potential similarities or associations between TreMs has potential theoretical value, e.g. it may help researchers identify common factors favouring TreM formation or help managers select trees with multiple TreMs as candidates for retention.
Accurate heritability estimates for fitness‐related traits are required to predict an organism’s ability to respond to global change. Heritability estimates are theoretically expected to be inflated ...if, due to limited dispersal, individuals that share genes are also likely to share similar environments. However, if relatives occupy similar environments due, at least partly, to genetic variation for habitat selection, then accounting for environmental similarity in quantitative genetic models may result in diminished heritability estimates in wild populations. This potential issue has been pointed out in the literature, but has not been evaluated by empirical studies.
Here, we investigate whether environmental similarity among individuals can be partly explained by genetic variation for habitat selection, and how this link potentially blurs estimates for heritability in fitness‐related traits.
Using intensive GPS monitoring, we quantified home‐range habitat composition for 293 roe deer inhabiting a heterogeneous landscape to assess environmental similarity. To investigate if environmental similarity might harbour genetic variation, we combined genome‐wide data in a quantitative genetic framework to evaluate genetic variation for home‐range habitat composition, which is partly the result of habitat selection at settlement. Finally, we explored how environmental similarity affects heritability estimates for behaviours related to the risk avoidance–resource acquisition trade‐off (i.e. being in open habitat and distance to roads) and proxies of individual performance (i.e. body mass and hind foot length). We found substantial heritability for home‐range habitat composition, with estimates ranging from 0.40 (proportion of meadows) to 0.85 (proportion of refuge habitat). Accounting for similarity in habitat composition between relatives decreased the heritability estimates for both behavioural and morphological traits (reduction ranging from 55% to 100% and from 22% to 41% respectively). As a consequence, only half of these heritability estimates remained significantly different from zero.
Our results show that similar genotypes occupy similar environments, which could lead to heritable variation being incorrectly attributed to environmental effects. To accurately distinguish the sources of phenotypic variation and predict the ability of organisms to respond to global change, it is necessary to develop quantitative genetic studies investigating the mechanisms underpinning environmental similarity among relatives.
Environmental similarity among individuals can, in part, be explained by genetic variation for habitat selection. Consequently, accounting for environmental similarity in quantitative genetics may blur the heritability estimates of fitness‐related traits by removing genuine genetic variance for a given trait.
The balance between resource acquisition and risk avoidance should vary according to personality type, with potential knock-on effects for fitness. Although previous studies have suggested a link ...between boldness and fitness components, little evidence is available on the behavioural mechanisms mediating this relationship in the wild. Because habitat use is the outcome of the trade-off between the costs and benefits associated with using each habitat type, we evaluated between-individual differences in habitat use of 64 GPS-collared female roe deer, Capreolus capreolus, using multinomial logit mixed models. To investigate whether deer differed in their habitat use tactics in relation to their personality type and their annual reproductive success, we assessed the link between individual habitat use patterns, boldness (measured as the strength of behavioural responsiveness to handling) and annual reproductive success (measured by the presence/absence of fawns at heel during autumn). Although daily and seasonal variations in the risk–resource landscapes clearly drove patterns of habitat use, individuals adopted contrasting habitat use tactics depending on their position along the shy–bold gradient and their reproductive status. Shy individuals occupied safer woodland more frequently, even at night when risk is lower. In contrast, bold individuals were better able to exploit rich open habitats. When this included mature autumn crops, these females weaned more offspring. Finally, irrespective of personality type, females that used meadows more often also achieved higher annual reproductive success. Overall, we demonstrate that individuals express divergent habitat use tactics as a function of their ability to avoid exposure to risk and their annual reproductive success.
•We measured habitat use, boldness and reproductive success in 64 female roe deer.•Spatiotemporal variations in the risk-resource trade-off shape deer habitat use.•The way individuals use their environment also depends on their personality type.•Differences in habitat use between personalities are linked to reproductive success.•Individual habitat use tactics mediate the personality – fitness link in the wild.
Quantifying the impact of land-use changes on biodiversity is a major challenge in conservation ecology. Static spatial relationships between bird communities and agricultural landscapes have been ...extensively studied. Yet, their ability to mirror the effects of temporal land-use dynamics remains to be demonstrated. Here, we test whether such space-for-time substitution approaches are relevant for explaining temporal variations in farmland bird communities. We surveyed 256 bird communities in an agricultural landscape in southwest France at the same locations in 1982 and 2007, and quantified the same seven landscape descriptors for each period. We compared the effects of spatial and temporal landscape changes over this 25-year period on bird species distributions and three community-level metrics: species richness and two community indices reflecting birds' specialisation regarding local vegetation structure (local CSI) and landscape composition (landscape CSI). Landscape heterogeneity decreased between 1982 and 2007 and crop area increased sharply at the expense of grassland as a result of agricultural intensification. We found that the correlations between temporal changes in bird distributions or community metrics and landscape components were less consistent than their spatial relationships in each year. This result advocates caution when using a space-for-time substitution approach to assess the effects of landscape changes on biodiversity. Additionally, community metrics showed contrasted responses to landscape changes. Species richness and local CSI for each period were negatively related to the area of crops and positively related to landscape heterogeneity. Conversely, the landscape CSI was positively related to the area of crop and negatively to landscape heterogeneity. To understand the ecological processes linked to changes in farm landscapes, our study underlines the need to develop long-term studies with bird and habitat data collected during several periods, and particularly to consider multiple community indices in monitoring change.
•The Pyrenean chamois, a near-monomorphic ruminant shows a marked sexual segregation.•Habitat segregation peaks in spring as expected with exclusive maternal care.•Social and spatial segregation was ...very high year-long and consistent across years.•Current hypotheses based on sexual dimorphism fail to account for the sexual segregation in this species.
Adult females and males live apart outside the mating period in many social vertebrates, but the causes of this phenomenon remain a matter of debate. Current prevailing hypotheses predict no sexual segregation outside the early period of maternal care in nearly monomorphic species such as the Pyrenean chamois (Rupicapra pyrenaica). We examined sexual segregation in a population of the species, using data collected over 143 consecutive months on groups’ location and composition, and extending statistical procedures introduced by Conradt (1998b) and Bonenfant et al. (2007). In addition, we analysed the social interactions recorded between group members. As expected, habitat segregation was low throughout the year, with a maximum during the early lactation period. However, social and spatial segregation was consistently high, contradicting the predictions of the current prevailing hypotheses, while suggesting social causes were predominant. The scarcity of social interactions outside the mating season makes unlikely the hypothesis that males segregate to improve their reproductive success. We rather suspect that higher social affinities within than between the two sexes are at work. However, this hypothesis alone is probably insufficient to account for spatial segregation. Our results should revive the debate regarding the causes of sexual segregation.