Biomechanical properties of non-flight vibrations produced by bees.

Bees use thoracic vibrations produced by their indirect flight muscles for powering wingbeats in flight, but also during mating, pollination, defence and nest building. Previous work on non-flight vibrations has mostly focused on acoustic (airborne vibrations) and spectral properties (frequency domain). However, mechanical properties such as the vibration's acceleration amplitude are important in some behaviours, e.g. during buzz pollination, where higher amplitude vibrations remove more pollen from flowers. Bee vibrations have been studied in only a handful of species and we know very little about how they vary among species. In this study, we conducted the largest survey to date of the biomechanical properties of non-flight bee buzzes. We focused on defence buzzes as they can be induced experimentally and provide a common currency to compare among taxa. We analysed 15,000 buzzes produced by 306 individuals in 65 species and six families from Mexico, Scotland and Australia. We found a strong association between body size and the acceleration amplitude of bee buzzes. Comparison of genera that buzz-pollinate and those that do not suggests that buzz-pollinating bees produce vibrations with higher acceleration amplitude. We found no relationship between bee size and the fundamental frequency of defence buzzes. Although our results suggest that body size is a major determinant of the amplitude of non-flight vibrations, we also observed considerable variation in vibration properties among bees of equivalent size and even within individuals. Both morphology and behaviour thus affect the biomechanical properties of non-flight buzzes.

Climatic and soil characteristics account for the genetic structure of the invasive cactus moth Cactoblastis cactorum, in its native range in Argentina.

Background: Knowledge of the physical and environmental conditions that may limit the migration of invasive species is crucial to assess the potential for expansion outside their native ranges. The cactus moth, Cactoblastis cactorum, is native to South America (Argentina, Paraguay, Uruguay and Brazil) and has been introduced and invaded the Caribbean and southern United States, among other regions. In North America there is an ongoing process of range expansion threatening cacti biodiversity of the genus Opuntia and the commercial profits of domesticated Opuntia ficus-indica. Methods: To further understand what influences the distribution and genetic structure of this otherwise important threat to native and managed ecosystems, in the present study we combined ecological niche modeling and population genetic analyses to identify potential environmental barriers in the native region of Argentina. Samples were collected on the host with the wider distribution range, O. ficus-indica. Results: Significant genetic structure was detected using 10 nuclear microsatellites and 24 sampling sites. At least six genetic groups delimited by mountain ranges, salt flats and wetlands were mainly located to the west of the Dry Chaco ecoregion. Niche modeling supports that this region has high environmental suitability where the upper soil temperature and humidity, soil carbon content and precipitation were the main environmental factors that explain the presence of the moth. Environmental filters such as the upper soil layer may be critical for pupal survival and consequently for the establishment of populations in new habitats, whereas the presence of available hosts is a necessary conditions for insect survival, upper soil and climatic characteristics will determine the opportunities for a successful establishment.

The role of within-plant variation in nectar production: an experimental approach.

BACKGROUND AND AIMS: Nectar, a plant reward for pollinators, can be energetically expensive. Hence, a higher investment in nectar production can lead to reduced allocation to other vital functions and/or increased geitonogamous pollination. One possible strategy employed by plants to reduce these costs is to offer variable amounts of nectar among flowers within a plant, to manipulate pollinator behaviour. Using artificial flowers, we tested this hypothesis by examining how pollinator visitation responds to inter- and intra-plant variation in nectar production, assessing how these responses impact the energetic cost per visit. METHODS: We conducted a 2 × 2 factorial experiment using artificial flowers, with two levels of nectar investment (high and low sugar concentration) and two degrees of intra-plant variation in nectar concentration (coefficient of variation 0 and 20 %). The experimental plants were exposed to visits (number and type) from a captive Bombus impatiens colony, and we recorded the total visitation rate, distinguishing geitonogamous from exogamous visits. Additionally, we calculated two estimators of the energetic cost per visit and examined whether flowers with higher nectar concentrations (richer flowers) attracted more bumblebees. KEY RESULTS: Plants in the variable nectar production treatment (coefficient of variation 20 %) had a greater proportion of flowers visited by pollinators, with higher rates of total, geitonogamous and exogamous visitation, compared with plants with invariable nectar production. When assuming no nectar reabsorption, variable plants incurred a lower cost per visit compared with invariable plants. Moreover, highly rewarding flowers on variable plants had higher rates of pollination visits compared with flowers with few rewards. CONCLUSIONS: Intra-plant variation in nectar concentration can represent a mechanism for pollinator manipulation, enabling plants to decrease the energetic costs of the interaction while still ensuring consistent pollinator visitation. However, our findings did not provide support for the hypothesis that intra-plant variation in nectar concentration acts as a mechanism to avoid geitonogamy. Additionally, our results confirmed the hypothesis that increased visitation to variable plants is dependent on the presence of flowers with nectar concentration above the mean.

Chained Deep Learning Using Generalized Cross-Entropy for Multiple Annotators Classification.

Supervised learning requires the accurate labeling of instances, usually provided by an expert. Crowdsourcing platforms offer a practical and cost-effective alternative for large datasets when individual annotation is impractical. In addition, these platforms gather labels from multiple labelers. Still, traditional multiple-annotator methods must account for the varying levels of expertise and the noise introduced by unreliable outputs, resulting in decreased performance. In addition, they assume a homogeneous behavior of the labelers across the input feature space, and independence constraints are imposed on outputs. We propose a Generalized Cross-Entropy-based framework using Chained Deep Learning (GCECDL) to code each annotator's non-stationary patterns regarding the input space while preserving the inter-dependencies among experts through a chained deep learning approach. Experimental results devoted to multiple-annotator classification tasks on several well-known datasets demonstrate that our GCECDL can achieve robust predictive properties, outperforming state-of-the-art algorithms by combining the power of deep learning with a noise-robust loss function to deal with noisy labels. Moreover, network self-regularization is achieved by estimating each labeler's reliability within the chained approach. Lastly, visual inspection and relevance analysis experiments are conducted to reveal the non-stationary coding of our method. In a nutshell, GCEDL weights reliable labelers as a function of each input sample and achieves suitable discrimination performance with preserved interpretability regarding each annotator's trustworthiness estimation.

Posthoc Interpretability of Neural Responses by Grouping Subject Motor Imagery Skills Using CNN-Based Connectivity.

Motor Imagery (MI) refers to imagining the mental representation of motor movements without overt motor activity, enhancing physical action execution and neural plasticity with potential applications in medical and professional fields like rehabilitation and education. Currently, the most promising approach for implementing the MI paradigm is the Brain-Computer Interface (BCI), which uses Electroencephalogram (EEG) sensors to detect brain activity. However, MI-BCI control depends on a synergy between user skills and EEG signal analysis. Thus, decoding brain neural responses recorded by scalp electrodes poses still challenging due to substantial limitations, such as non-stationarity and poor spatial resolution. Also, an estimated third of people need more skills to accurately perform MI tasks, leading to underperforming MI-BCI systems. As a strategy to deal with BCI-Inefficiency, this study identifies subjects with poor motor performance at the early stages of BCI training by assessing and interpreting the neural responses elicited by MI across the evaluated subject set. Using connectivity features extracted from class activation maps, we propose a Convolutional Neural Network-based framework for learning relevant information from high-dimensional dynamical data to distinguish between MI tasks while preserving the post-hoc interpretability of neural responses. Two approaches deal with inter/intra-subject variability of MI EEG data: (a) Extracting functional connectivity from spatiotemporal class activation maps through a novel kernel-based cross-spectral distribution estimator, (b) Clustering the subjects according to their achieved classifier accuracy, aiming to find common and discriminative patterns of motor skills. According to the validation results obtained on a bi-class database, an average accuracy enhancement of 10% is achieved compared to the baseline EEGNet approach, reducing the number of "poor skill" subjects from 40% to 20%. Overall, the proposed method can be used to help explain brain neural responses even in subjects with deficient MI skills, who have neural responses with high variability and poor EEG-BCI performance.

Geographic isolation and long-distance gene flow influence the genetic structure of the blue fan palm Brahea armata (Arecaceae).

The formation of the Baja California Peninsula (BCP) has impacted the microevolutionary dynamics of different species in ways that depend on biological traits such as dispersal capacity. Plants with relatively low levels of vagility have exhibited high genetic divergence between the BCP and Continental mainland. Brahea armata (Arecaceae) is a palm species inhabiting the northern part of the BCP and Sonora; its distribution occurs in isolated oases of vegetation. We aimed to evaluate the influence of the formation of the BCP on the genetic structure of B. armata using nuclear microsatellites and chloroplast markers (cpDNA) to compare patterns of genetic diversity and structure with previous published studies. Because gene flow through seeds is usually more limited compared to pollen flow, we expect to find stronger genetic structure at (cpDNA) than at nuclear markers. Moreover, larger genetic structure might also be explained by the smaller effective population size of cpDNA. We analyzed six microsatellite markers and two cpDNA regions. The main results indicated high levels of genetic differentiation among isolated populations located in the BCP, while low genetic differentiation was found between southern populations of the BCP and Sonora, suggesting long distance gene flow. In contrast, chloroplast markers indicated high levels of genetic structure between BCP and Sonora populations, suggesting asymmetrical gene flow between pollen (measured by nuclear microsatellites) and seed (cpDNA markers). This study provides valuable information on genetic diversity of B. armata that can be relevant for conservation and management; and develops microsatellites markers that can be transferred to other Brahea species.

KCS-FCnet: Kernel Cross-Spectral Functional Connectivity Network for EEG-Based Motor Imagery Classification.

This paper uses EEG data to introduce an approach for classifying right and left-hand classes in Motor Imagery (MI) tasks. The Kernel Cross-Spectral Functional Connectivity Network (KCS-FCnet) method addresses these limitations by providing richer spatial-temporal-spectral feature maps, a simpler architecture, and a more interpretable approach for EEG-driven MI discrimination. In particular, KCS-FCnet uses a single 1D-convolutional-based neural network to extract temporal-frequency features from raw EEG data and a cross-spectral Gaussian kernel connectivity layer to model channel functional relationships. As a result, the functional connectivity feature map reduces the number of parameters, improving interpretability by extracting meaningful patterns related to MI tasks. These patterns can be adapted to the subject's unique characteristics. The validation results prove that introducing KCS-FCnet shallow architecture is a promising approach for EEG-based MI classification with the potential for real-world use in brain-computer interface systems.

Binary and multi-class automated detection of age-related macular degeneration using convolutional- and transformer-based architectures.

BACKGROUND AND OBJECTIVE: Age-related macular degeneration (AMD) is an eye disease that happens when ageing causes damage to the macula, and it is the leading cause of blindness in developed countries. Screening retinal fundus images allows ophthalmologists to early detect, diagnose and treat this disease; however, the manual interpretation of images is a time-consuming task. In this paper, we aim to study different deep learning methods to diagnose AMD. METHODS: We have conducted a thorough study of two families of deep learning models based on convolutional neural networks (CNN) and transformer architectures to automatically diagnose referable/non-referable AMD, and grade AMD severity scales (no AMD, early AMD, intermediate AMD, and advanced AMD). In addition, we have analysed several progressive resizing strategies and ensemble methods for convolutional-based architectures to further improve the performance of the models. RESULTS: As a first result, we have shown that transformer-based architectures obtain considerably worse results than convolutional-based architectures for diagnosing AMD. Moreover, we have built a model for diagnosing referable AMD that yielded a mean F1-score (SD) of 92.60% (0.47), a mean AUROC (SD) of 97.53% (0.40), and a mean weighted kappa coefficient (SD) of 85.28% (0.91); and an ensemble of models for grading AMD severity scales with a mean accuracy (SD) of 82.55% (2.92), and a mean weighted kappa coefficient (SD) of 84.76% (2.45). CONCLUSIONS: This work shows that working with convolutional based architectures is more suitable than using transformer based models for classifying and grading AMD from retinal fundus images. Furthermore, convolutional models can be improved by means of progressive resizing strategies and ensemble methods.

Priming by Insects: Differential Effects of Sympatric and Allopatric Priming upon Plant Performance and Tolerance to Herbivory.

Plants have evolved multiple mechanisms to defend themselves from their multiple herbivores. Thus, being able to recognise among them and respond accordingly is fundamental for plant survival and reproduction. Defence priming prepares the plant to better or more rapidly respond to future damage; however, while it is considered an adaptive trait, to date, no studies have evaluated the extent and specificity of the priming recognition. To estimate the costs, benefits and specificity of priming, we used a highly specialist plant-insect system (Datura stramonium-Lema daturaphila) and performed a reciprocal transplant experiment with two populations where a priming stimulus (sympatric vs. allopatric) and a damage treatment (sympatric) were applied. We found no evidence of a fitness cost of priming, given that primed plants without damage showed no reduction in fitness. In contrast, our treatments affected the probability of bud abortion. That is, when damaged plants received no priming or the priming came from an allopatric insect, the likelihood of aborting the first bud was 1.9 times greater compared to plants being primed by their sympatric insect. We also found that damaged plants primed with an allopatric insect produced 14% fewer seeds compared to plants receiving a sympatric priming stimulus. Tolerance to herbivore damage was also the lowest when plants received the priming stimulus from an allopatric insect. Overall, these results suggest that, in our study system, plants recognise their local insect population reducing the negative effect of damage through a tolerance response.

Combined Immunoglobulin Free Light Chains Are Novel Predictors of Cardiovascular Events in Patients With Abdominal Aortic Aneurysm.

OBJECTIVE: Abdominal aortic aneurysm (AAA) is characterised by the presence of B cells and immunoglobulins in the aortic wall, mainly in the adventitia. Kappa (κ) and lambda (λ) free light chains (FLCs) are produced from B cells during immunoglobulin synthesis. This study investigated the presence and prognostic value of combined FLCs (cFLCs or summed κ and λ) in patients with AAA. METHODS: cFLCs were analysed by a turbidimetric specific assay in tissue conditioned media from AAA samples (n = 34) compared with healthy aortas (n = 34) from France and in plasma samples from patients with AAA (n = 434) and age matched controls (n = 104) selected from the Viborg Vascular (VIVA) AAA screening trial in Denmark. t test, logistic regression, and Cox regression were used to test whether plasma cFLCs serve as a marker for AAA presence and whether cFLCs were predictive of death, major adverse cardiovascular events (MACE), or major adverse lower limb events (MALE). RESULTS: Increased cFLC levels were detected in the AAA adventitial layer compared with the AAA medial layer and healthy media layer (13.65 ± 3.17 vs. 6.57 ± 1.01 vs. 0.49 ± 0.09 mg/L, respectively, p < .050). The upper tertile of plasma cFLCs was independently associated with AAA presence after correcting for confounders (odds ratio [OR] 7.596, 95% confidence intervals [CI] 3.117 - 18.513; p < .001). Of 434 patients with AAA, 89 (20.5%) died, 104 (24.0%) suffered MACE, and 63 (14.5%) suffered MALE, during a five year follow up. In univariable analysis, the cFLC upper tertile was associated with a higher risk of death, MACE, and MALE (p < .001 for all). After adjustment for confounders, cFLCs remained an independent predictor of all cause mortality (hazard ratio [HR] 4.310, 95% CI 2.157 - 8.609; p < .001), MACE (HR 2.153, 95% CI 1.218 - 3.804; p = .008), or MALE (HR 3.442, 95% CI 1.548 - 7.652; p = .002) for those in the upper tertile. CONCLUSION: Increased cFLCs are observed in adventitial tissue of patients with AAA, indicating local activation of B cells. Plasma cFLC levels are an independent predictor of death, MACE, and MALE in patients with AAA.

Roll-to-roll nanoimprint lithography of high efficiency Fresnel lenses for micro-concentrator photovoltaics.

Roll-to-roll nanoimprint lithography (R2R-NIL) is an enabling technology for the low-cost mass production of high-quality micro- and nano-sized optical elements. Particularly, the fabrication of Fresnel lenses using R2R-NIL is a promising approach to produce optical arrays for micro-concentrator photovoltaic modules. This work investigates the application of a continuous R2R imprinting process based on ultraviolet curing of transparent photopolymer resins (UV-NIL) to fabricate high-efficiency and low-cost Fresnel lenses. The morphological attributes and the related optical performance of the lenses fabricated using roll-to-roll UV-NIL on flexible PET sheets yielded optical efficiency values up to ∼ 69% at a concentration ratio of 178X, whereas a value of ∼ 77% was obtained for the UV-NIL batch processed on a flat rigid substrate. Further improvement of the optical efficiency has been achieved by adding moth-eye inspired antireflective (AR) features on the side opposite to the Fresnel motifs via a double-sided R2R UV-NIL process. The process developed paves the way for cost-effective mass production of high-efficiency Fresnel lenses for micro-concentrator photovoltaics.

Comparison of achromatic doublet on glass Fresnel lenses for concentrator photovoltaics.

Silicone on glass (SoG) Fresnel lenses are the reference technology in concentrator photovoltaics (CPV) because of their simplicity and low cost. Nevertheless, their performance is strongly limited by chromatic aberration. As an alternative, in order to overcome such limitation, achromatic doublet on glass (ADG) Fresnel lenses were proposed. Such lenses are achromatic cemented doublet specifically designed for CPV applications. In this paper, a novel ADG architecture is presented and its performance analyzed and compared to previous proposals. The results show that most of the intrinsic optical losses are minimized and a superior optical efficiency can be achieved. The novel ADG design provides an achromatic lens for CPV whose efficiency is almost equal to the reference SoG technology and, at the same time, maintains all the advantages provided by the achromatic design such as the higher maximum attainable concentration and the strongly reduced temperature dependency.

MotilityJ: An open-source tool for the classification and segmentation of bacteria on motility images.

BACKGROUND AND OBJECTIVES: Infectious diseases produced by antimicrobial resistant microorganisms are a major threat to human, and animal health worldwide. This problem is increased by the virulence and spread of these bacteria. Surface motility has been regarded as a pathogenicity element because it is essential for many biological functions, but also for disease spreading; hence, investigations on the motility behaviour of bacteria are crucial to understand chemotaxis, biofilm formation and virulence in general. To identify a motile strain in the laboratory, the bacterial spread area is observed on media solidified with agar. Up to now, the task of measuring bacteria spread was a manual, and, therefore, tedious and time-consuming task. The aim of this work is the development of a set of tools for bacteria segmentation in motility images. METHODS: In this work, we address the problem of measuring bacteria spread on motility images by creating an automatic pipeline based on deep learning models. Such a pipeline consists of a classification model to determine whether the bacteria has spread to cover completely the Petri dish, and a segmentation model to determine the spread of those bacteria that do not fully cover the Petri dishes. In order to annotate enough images to train our deep learning models, a semi-automatic annotation procedure is presented. RESULTS: The classification model of our pipeline achieved a F1-score of 99.85%, and the segmentation model achieved a Dice coefficient of 95.66%. In addition, the segmentation model produces results that are indistinguishable, and in many cases preferred, from those produced manually by experts. Finally, we facilitate the dissemination of our pipeline with the development of MotilityJ, an open-source and user-friendly application for measuring bacteria spread on motility images. CONCLUSIONS: In this work, we have developed an algorithm and trained several models for measuring bacteria spread on motility images. Thanks to this work, the analysis of motility images will be faster and more reliable. The developed tools will help to advance our understanding of the behaviour and virulence of bacteria.

SpheroidJ: An Open-Source Set of Tools for Spheroid Segmentation.

BACKGROUND AND OBJECTIVES: Spheroids are the most widely used 3D models for studying the effects of different micro-environmental characteristics on tumour behaviour, and for testing different preclinical and clinical treatments. In order to speed up the study of spheroids, imaging methods that automatically segment and measure spheroids are instrumental; and, several approaches for automatic segmentation of spheroid images exist in the literature. However, those methods fail to generalise to a diversity of experimental conditions. The aim of this work is the development of a set of tools for spheroid segmentation that works in a diversity of settings. METHODS: In this work, we have tackled the spheroid segmentation task by first developing a generic segmentation algorithm that can be easily adapted to different scenarios. This generic algorithm has been employed to reduce the burden of annotating a dataset of images that, in turn, has been employed to train several deep learning architectures for semantic segmentation. Both our generic algorithm and the constructed deep learning models have been tested with several datasets of spheroid images where the spheroids were grown under several experimental conditions, and the images acquired using different equipment. RESULTS: The developed generic algorithm can be particularised to different scenarios; however, those particular algorithms fail to generalise to different conditions. By contrast, the best deep learning model, constructed using the HRNet-Seg architecture, generalises properly to a diversity of scenarios. In order to facilitate the dissemination and use of our algorithms and models, we present SpheroidJ, a set of open-source tools for spheroid segmentation. CONCLUSIONS: In this work, we have developed an algorithm and trained several models for spheroid segmentation that can be employed with images acquired under different conditions. Thanks to this work, the analysis of spheroids acquired under different conditions will be more reliable and comparable; and, the developed tools will help to advance our understanding of tumour behaviour.

Effects of herbivory and its timing on reproductive success of a tropical deciduous tree.

BACKGROUND AND AIMS: The implications of herbivory for plant reproduction have been widely studied; however, the relationship of defoliation and reproductive success is not linear, as there are many interacting factors that may influence reproductive responses to herbivore damage. In this study we aimed to disentangle how the timing of foliar damage impacts both male and female components of fitness, and to assess when it has greater impacts on plant reproductive success. METHODS: We measured herbivore damage and its effects on floral production, male and female floral attributes as well as fruit yield in three different phenological phases of Casearia nitida (Salicaceae) over the course of two consecutive years. Then we tested two models of multiple causal links among herbivory and reproductive success using piecewise structural equation models. KEY RESULTS: The effects of leaf damage differed between reproductive seasons and between male and female components of fitness. Moreover, the impact of herbivory extended beyond the year when it was exerted. The previous season's cumulated foliar damage had the largest impact on reproductive characters, in particular a negative effect on the numbers of inflorescences, flowers and pollen grains, indirectly affecting the numbers of infructescences and fruits, and a positive one on the amount of foliar damage during flowering. CONCLUSIONS: For perennial and proleptic species, the dynamics of resource acquisition and allocation patterns for reproduction promote and extend the effects of herbivore damage to longer periods than a single reproductive event and growing season, through the interactions among different components of female and male fitness.

Disentangling the effects of local resources, landscape heterogeneity and climatic seasonality on bee diversity and plant-pollinator networks in tropical highlands.

Land-use alteration and climate seasonality have profound effects on bee species diversity by influencing the availability of nesting and floral resources. Here, using twelve sites embedded in an agriculture-forest mosaic in the tropical highlands of Guatemala, we investigated the relative effects of climate seasonality and landscape heterogeneity on bee and floral-resource community structure and on their mutualistic network architecture. We found that climate seasonality affected bee diversity, which was higher in the wet season and associated positively with the availability of floral resources across both seasons. Bee community composition also differed between seasons and it was mainly driven by floral-resource richness and the proportion of agricultural, semi-natural and forest cover. In addition to the effects on bee diversity, climate seasonality also affected flower-bee visitation networks. We documented higher relative (null model corrected) nestedness in the dry season compared to the wet season. Niche partitioning as a result of competition for scarce resources in the dry season could be the process driving the differences in the network structure between seasons. Furthermore, relative nestedness was consistently smaller than zero, and relative modularity and specialization were consistently larger than zero in both seasons, suggesting the existence of isolated groups of interacting partners in all our flower-bee visitation networks. Our results highlight the effect of climatic seasonality and the importance of preserving local floral resources and natural heterogeneous habitats for the conservation of bee communities and their pollination services in tropical highlands.

Local dispersal pathways during the invasion of the cactus moth, Cactoblastis cactorum, within North America and the Caribbean.

Cactoblastis cactorum, a species of moth native to Argentina, feeds on several prickly pear cactus species (Opuntia) and has been successfully used as a biological control of invading Opuntia species in Australia, South Africa and native ruderal Opuntia species in some Caribbean islands. Since its introduction to the Caribbean its spread was uncontrolled, invading successfully Florida, Texas and Louisiana. Despite this long history of invasion, we are still far from understanding the factors determining the patterns of invasion of Cactoblastis in North America. Here, we explored three non-mutually exclusive explanations: a) a stepping stone model of colonization, b) long distance colonization due to hurricanes, and/or c) hitchhiking through previously reported commercial routes. Genetic diversity, genetic structure and the patterns of migration among populations were obtained by analyzing 10 nuclear microsatellite loci. Results revealed the presence of genetic structure among populations of C. cactorum in the invaded region and suggest that both marine commercial trade between the Caribbean islands and continental USA, as well as recurrent transport by hurricanes, explain the observed patterns of colonization. Provided that sanitary regulations avoiding human-mediated dispersal are enforced, hurricanes probably represent the most important agent of dispersal and future invasion to continental areas.

Ontogenetic changes in the targets of natural selection in three plant defenses.

The evolution of plant defenses has traditionally been studied at single plant ontogenetic stages, overlooking the fact that natural selection acts continuously on organisms along their development, and that the adaptive value of phenotypes can change along ontogeny. We exposed 20 replicated genotypes of Turnera velutina to field conditions to evaluate whether the targets of natural selection on different defenses and their adaptative value change across plant development. We found that low chemical defense was favored in seedlings, which seems to be explained by the assimilation efficiency and the ability of the specialist herbivore to sequester cyanogenic glycosides. Whereas trichome density was unfavored in juvenile plants, it increased relative plant fitness in reproductive plants. At this stage we also found a positive correlative gradient between cyanogenic potential and sugar content in extrafloral nectar. We visualize this complex multi-trait combination as an ontogenetic defensive strategy. The inclusion of whole-plant ontogeny as a key source of variation in plant defense revealed that the targets and intensity of selection change along the development of plants, indicating that the influence of natural selection cannot be inferred without the assessment of ontogenetic strategies in the expression of multiple defenses.

Natural selection acting on integrated phenotypes: covariance among functional leaf traits increases plant fitness.

Plant functional strategies are usually accomplished through the simultaneous expression of different traits, and hence their correlations should be promoted by natural selection. The adaptive value of correlations among leaf functional traits, however, has not been assessed in natural populations. We estimated intraspecific variation in leaf functional traits related to the primary metabolism and anti-herbivore defence in a population of Turnera velutina. We analysed whether natural selection favoured the expression of individual traits, particular combinations of traits or leaf phenotypic integration. Patterns of covariation among traits were related to water and nitrogen economy, and were similar among genotypes, but the magnitude of their phenotypic integration differed by 10-fold. Although families did not differ in the mean values of leaf functional traits, directional selection favoured low nitrogen content and low chemical defence, high content of chlorophyll, sugar in extrafloral nectar and trichome density. Families with higher phenotypic integration among leaf traits grew faster and produced more flowers. We suggest that the coordinated expression of leaf traits has an adaptive value, probably related to optimisation in the expression of traits related to water conservation and nitrogen acquisition.