Cross-biome synthesis of source versus sink limits to tree growth.

Uncertainties surrounding tree carbon allocation to growth are a major limitation to projections of forest carbon sequestration and response to climate change. The prevalence and extent to which carbon assimilation (source) or cambial activity (sink) mediate wood production are fundamentally important and remain elusive. We quantified source-sink relations across biomes by combining eddy-covariance gross primary production with extensive on-site and regional tree ring observations. We found widespread temporal decoupling between carbon assimilation and tree growth, underpinned by contrasting climatic sensitivities of these two processes. Substantial differences in assimilation-growth decoupling between angiosperms and gymnosperms were determined, as well as stronger decoupling with canopy closure, aridity, and decreasing temperatures. Our results reveal pervasive sink control over tree growth that is likely to be increasingly prominent under global climate change.

Thermodynamic Driving Forces for Substrate Atom Extraction by Adsorption of Strong Electron Acceptor Molecules.

A quantitative structural investigation is reported, aimed at resolving the issue of whether substrate adatoms are incorporated into the monolayers formed by strong molecular electron acceptors deposited onto metallic electrodes. A combination of normal-incidence X-ray standing waves, low-energy electron diffraction, scanning tunnelling microscopy, and X-ray photoelectron spectroscopy measurements demonstrate that the systems TCNQ and F4TCNQ on Ag(100) lie at the boundary between these two possibilities and thus represent ideal model systems with which to study this effect. A room-temperature commensurate phase of adsorbed TCNQ is found not to involve Ag adatoms, but to adopt an inverted bowl configuration, long predicted but not previously identified experimentally. By contrast, a similar phase of adsorbed F4TCNQ does lead to Ag adatom incorporation in the overlayer, the cyano end groups of the molecule being twisted relative to the planar quinoid ring. Density functional theory (DFT) calculations show that this behavior is consistent with the adsorption energetics. Annealing of the commensurate TCNQ overlayer phase leads to an incommensurate phase that does appear to incorporate Ag adatoms. Our results indicate that the inclusion (or exclusion) of metal atoms into the organic monolayers is the result of both thermodynamic and kinetic factors.

The complexity of biological disturbance agents, fuels heterogeneity, and fire in coniferous forests of the western United States.

Forest biological disturbance agents (BDAs) are insects, pathogens, and parasitic plants that affect tree decline, mortality, and forest ecosystems processes. BDAs are commonly thought to increase the likelihood and severity of fire by converting live standing trees to more flammable, dead and downed fuel. However, recent research indicates that BDAs do not necessarily increase, and can reduce, the likelihood or severity of fire. This has led to confusion regarding the role of BDAs in influencing fuels and fire in fire-prone western United States forests. Here, we review the existing literature on BDAs and their effects on fuels and fire in the western US and develop a conceptual framework to better understand the complex relationships between BDAs, fuels and fire. We ask: 1) What are the major BDA groups in western US forests that affect fuels? and 2) How do BDA-affected fuels influence fire risk and outcomes? The conceptual framework is rooted in the spatiotemporal aspects of BDA life histories, which drive forest impacts, fuel characteristics and if ignited, fire outcomes. Life histories vary among BDAs from episodic, landscape-scale outbreaks (bark beetles, defoliators), to chronic, localized disturbance effects (dwarf mistletoes, root rots). Generally, BDAs convert aboveground live biomass to dead biomass, decreasing canopy fuels and increasing surface fuels. However, the rate of conversion varies with time-since-event and among BDAs and forest types, resulting in a wide range of effects on the amount of dead fuels at any given time and place, which interacts with the structure and composition of the stand before and subsequent to BDA events. A major influence on fuels may be that BDAs have emerged as dominant agents of forest heterogeneity creation. Because BDAs play complex roles in fuels and fire heterogeneity across the western US which are further complicated by interactions with climate change, drought, and forest management (fire suppression), their impacts on fuels, fire and ecological consequences cannot be categorized simply as positive or negative but need to be evaluated within the context of BDA life histories and ecosystem dynamics.

A Fast, Provably Accurate Approximation Algorithm for Sparse Principal Component Analysis Reveals Human Genetic Variation Across the World.

Principal component analysis (PCA) is a widely used dimensionality reduction technique in machine learning and multivariate statistics. To improve the interpretability of PCA, various approaches to obtain sparse principal direction loadings have been proposed, which are termed Sparse Principal Component Analysis (SPCA). In this paper, we present ThreSPCA, a provably accurate algorithm based on thresholding the Singular Value Decomposition for the SPCA problem, without imposing any restrictive assumptions on the input covariance matrix. Our thresholding algorithm is conceptually simple; much faster than current state-of-the-art; and performs well in practice. When applied to genotype data from the 1000 Genomes Project, ThreSPCA is faster than previous benchmarks, at least as accurate, and leads to a set of interpretable biomarkers, revealing genetic diversity across the world.

Hutch++: Optimal Stochastic Trace Estimation.

We study the problem of estimating the trace of a matrix A that can only be accessed through matrix-vector multiplication. We introduce a new randomized algorithm, Hutch++, which computes a (1 ± ε) approximation to tr( A ) for any positive semidefinite (PSD) A using just O(1/ε) matrix-vector products. This improves on the ubiquitous Hutchinson's estimator, which requires O(1/ε 2) matrix-vector products. Our approach is based on a simple technique for reducing the variance of Hutchinson's estimator using a low-rank approximation step, and is easy to implement and analyze. Moreover, we prove that, up to a logarithmic factor, the complexity of Hutch++ is optimal amongst all matrix-vector query algorithms, even when queries can be chosen adaptively. We show that it significantly outperforms Hutchinson's method in experiments. While our theory requires A to be positive semidefinite, empirical gains extend to applications involving non-PSD matrices, such as triangle estimation in networks.

Subspace exploration: Bounds on Projected Frequency Estimation.

Given an n × d dimensional dataset A, a projection query specifies a subset C ⊆ [d] of columns which yields a new n × |C| array. We study the space complexity of computing data analysis functions over such subspaces, including heavy hitters and norms, when the subspaces are revealed only after observing the data. We show that this important class of problems is typically hard: for many problems, we show 2Ω(d) lower bounds. However, we present upper bounds which demonstrate space dependency better than 2 d . That is, for c, c' ∈ (0, 1) and a parameter N = 2 d an Nc -approximation can be obtained in space min ( N c ' , n ) , showing that it is possible to improve on the naïve approach of keeping information for all 2 d subsets of d columns. Our results are based on careful constructions of instances using coding theory and novel combinatorial reductions that exhibit such space-approximation tradeoffs.

An Ecological Perspective on Living with Fire in Ponderosa Pine Forests of Oregon and Washington: Resistance, Gone but not Forgotten.

Wildland fires (WLF) have become more frequent, larger, and severe with greater impacts to society and ecosystems and dramatic increases in firefighting costs. Forests throughout the range of ponderosa pine in Oregon and Washington are jeopardized by the interaction of anomalously dense forest structure, a warming and drying climate, and an expanding human population. These forests evolved with frequent interacting disturbances including low-severity surface fires, droughts, and biological disturbance agents (BDAs). Chronic low-severity disturbances were, and still are, critical to maintaining disturbance resistance, the property of an ecosystem to withstand disturbance while maintaining its structure and ecological function. Restoration of that historical resistance offers multiple social and ecological benefits. Moving forward, we need a shared understanding of the ecology of ponderosa pine forests to appreciate how restoring resistance can reduce the impacts of disturbances. Given contemporary forest conditions, a warming climate, and growing human populations, we predict continued elevation of tree mortality from drought, BDAs, and the large high-severity WLFs that threaten lives and property as well as ecosystem functions and services. We recommend more comprehensive planning to promote greater use of prescribed fire and management of reported fires for ecological benefits, plus increased responsibility and preparedness of local agencies, communities and individual homeowners for WLF and smoke events. Ultimately, by more effectively preparing for fire in the wildland urban interface, and by increasing the resistance of ponderosa pine forests, we can greatly enhance our ability to live with fire and other disturbances.

How to reduce dimension with PCA and random projections?

In our "big data" age, the size and complexity of data is steadily increasing. Methods for dimension reduction are ever more popular and useful. Two distinct types of dimension reduction are "data-oblivious" methods such as random projections and sketching, and "data-aware" methods such as principal component analysis (PCA). Both have their strengths, such as speed for random projections, and data-adaptivity for PCA. In this work, we study how to combine them to get the best of both. We study "sketch and solve" methods that take a random projection (or sketch) first, and compute PCA after. We compute the performance of several popular sketching methods (random iid projections, random sampling, subsampled Hadamard transform, CountSketch, etc) in a general "signal-plus-noise" (or spiked) data model. Compared to well-known works, our results (1) give asymptotically exact results, and (2) apply when the signal components are only slightly above the noise, but the projection dimension is non-negligible. We also study stronger signals allowing more general covariance structures. We find that (a) signal strength decreases under projection in a delicate way depending on the structure of the data and the sketching method, (b) orthogonal projections are slightly more accurate, (c) randomization does not hurt too much, due to concentration of measure, (d) CountSketch can be somewhat improved by a normalization method. Our results have implications for statistical learning and data analysis. We also illustrate that the results are highly accurate in simulations and in analyzing empirical data.

Alkali Doping Leads to Charge-Transfer Salt Formation in a Two-Dimensional Metal-Organic Framework.

Efficient charge transfer across metal-organic interfaces is a key physical process in modern organic electronics devices, and characterization of the energy level alignment at the interface is crucial to enable a rational device design. We show that the insertion of alkali atoms can significantly change the structure and electronic properties of a metal-organic interface. Coadsorption of tetracyanoquinodimethane (TCNQ) and potassium on a Ag(111) surface leads to the formation of a two-dimensional charge transfer salt, with properties quite different from those of the two-dimensional Ag adatom TCNQ metal-organic framework formed in the absence of K doping. We establish a highly accurate structural model by combination of quantitative X-ray standing wave measurements, scanning tunnelling microscopy, and density-functional theory (DFT) calculations. Full agreement between the experimental data and the computational prediction of the structure is only achieved by inclusion of a charge-transfer-scaled dispersion correction in the DFT, which correctly accounts for the effects of strong charge transfer on the atomic polarizability of potassium. The commensurate surface layer formed by TCNQ and K is dominated by strong charge transfer and ionic bonding and is accompanied by a structural and electronic decoupling from the underlying metal substrate. The consequence is a significant change in energy level alignment and work function compared to TCNQ on Ag(111). Possible implications of charge-transfer salt formation at metal-organic interfaces for organic thin-film devices are discussed.

Sketching algorithms for genomic data analysis and querying in a secure enclave.

Genome-wide association studies (GWAS), especially on rare diseases, may necessitate exchange of sensitive genomic data between multiple institutions. Since genomic data sharing is often infeasible due to privacy concerns, cryptographic methods, such as secure multiparty computation (SMC) protocols, have been developed with the aim of offering privacy-preserving collaborative GWAS. Unfortunately, the computational overhead of these methods remain prohibitive for human-genome-scale data. Here we introduce SkSES (https://github.com/ndokmai/sgx-genome-variants-search), a hardware-software hybrid approach for privacy-preserving collaborative GWAS, which improves the running time of the most advanced cryptographic protocols by two orders of magnitude. The SkSES approach is based on trusted execution environments (TEEs) offered by current-generation microprocessors-in particular, Intel's SGX. To overcome the severe memory limitation of the TEEs, SkSES employs novel 'sketching' algorithms that maintain essential statistical information on genomic variants in input VCF files. By additionally incorporating efficient data compression and population stratification reduction methods, SkSES identifies the top k genomic variants in a cohort quickly, accurately and in a privacy-preserving manner.

Near Optimal Linear Algebra in the Online and Sliding Window Models.

We initiate the study of numerical linear algebra in the sliding window model, where only the most recent W updates in a stream form the underlying data set. Although many existing algorithms in the sliding window model use or borrow elements from the smooth histogram framework (Braverman and Ostrovsky, FOCS 2007), we show that many interesting linear-algebraic problems, including spectral and vector induced matrix norms, generalized regression, and lowrank approximation, are not amenable to this approach in the row-arrival model. To overcome this challenge, we first introduce a unified row-sampling based framework that gives randomized algorithms for spectral approximation, low-rank approximation/projection-cost preservation, and ℓ 1-subspace embeddings in the sliding window model, which often use nearly optimal space and achieve nearly input sparsity runtime. Our algorithms are based on "reverse online" versions of offline sampling distributions such as (ridge) leverage scores, ℓ 1 sensitivities, and Lewis weights to quantify both the importance and the recency of a row; our structural results on these distributions may be of independent interest for future algorithmic design. Although our techniques initially address numerical linear algebra in the sliding window model, our row-sampling framework rather surprisingly implies connections to the well-studied online model; our structural results also give the first sample optimal (up to lower order terms) online algorithm for low-rank approximation/projection-cost preservation. Using this powerful primitive, we give online algorithms for column/row subset selection and principal component analysis that resolves the main open question of Bhaskara et al. (FOCS 2019). We also give the first online algorithm for ℓ 1-subspace embeddings. We further formalize the connection between the online model and the sliding window model by introducing an additional unified framework for deterministic algorithms using a merge and reduce paradigm and the concept of online coresets, which we define as a weighted subset of rows of the input matrix that can be used to compute a good approximation to some given function on all of its prefixes. Our sampling based algorithms in the row-arrival online model yield online coresets, giving deterministic algorithms for spectral approximation, low-rank approximation/projection-cost preservation, and ℓ 1-subspace embeddings in the sliding window model that use nearly optimal space.

Water potential control of turgor-driven tracheid enlargement in Scots pine at its xeric distribution edge.

The extent to which water availability can be used to predict the enlargement and final dimensions of xylem conduits remains an open issue. We reconstructed the time course of tracheid enlargement in Pinus sylvestris trees in central Spain by repeated measurements of tracheid diameter on microcores sampled weekly during a 2 yr period. We analyzed the role of water availability in these dynamics empirically through time-series correlation analysis and mechanistically by building a model that simulates daily tracheid enlargement rate and duration based on Lockhart's equation and water potential as the sole input. Tracheid enlargement followed a sigmoid-like time course, which varied intra- and interannually. Our empirical analysis showed that final tracheid diameter was strongly related to water availability during tracheid enlargement. The mechanistic model was calibrated and successfully validated (R2  = 0.92) against the observed tracheid enlargement time course. The model was also able to reproduce the seasonal variations of tracheid enlargement rate, duration and final diameter (R2  = 0.84-0.99). Our results support the hypothesis that tracheid enlargement and final dimensions can be modeled based on the direct effect of water potential on turgor-driven cell expansion. We argue that such a mechanism is consistent with other reported patterns of tracheid dimension variation.

Coordination and trade-offs between leaf and stem hydraulic traits and stomatal regulation along a spectrum of isohydry to anisohydry.

The degree of plant iso/anisohydry, a widely used framework for classifying species-specific hydraulic strategies, integrates multiple components of the whole-plant hydraulic pathway. However, little is known about how it associates with coordination of functional and structural traits within and across different organs. We examined stem and leaf hydraulic capacitance and conductivity/conductance, stem xylem anatomical features, stomatal regulation of daily minimum leaf and stem water potential (Ψ), and the kinetics of stomatal responses to vapour pressure deficit (VPD) in six diverse woody species differing markedly in their degree of iso/anisohydry. At the stem level, more anisohydric species had higher wood density and lower native capacitance and conductivity. Like stems, leaves of more anisohydric species had lower hydraulic conductance; however, unlike stems, their leaves had higher native capacitance at their daily minimum values of leaf Ψ. Moreover, rates of VPD-induced stomatal closure were related to intrinsic rather than native leaf capacitance and were not associated with species' degree of iso/anisohydry. Our results suggest a trade-off between hydraulic storage and efficiency in the leaf, but a coordination between hydraulic storage and efficiency in the stem along a spectrum of plant iso/anisohydry.

Co-occurring woody species have diverse hydraulic strategies and mortality rates during an extreme drought.

From 2011 to 2013, Texas experienced its worst drought in recorded history. This event provided a unique natural experiment to assess species-specific responses to extreme drought and mortality of four co-occurring woody species: Quercus fusiformis, Diospyros texana, Prosopis glandulosa, and Juniperus ashei. We examined hypothesized mechanisms that could promote these species' diverse mortality patterns using postdrought measurements on surviving trees coupled to retrospective process modelling. The species exhibited a wide range of gas exchange responses, hydraulic strategies, and mortality rates. Multiple proposed indices of mortality mechanisms were inconsistent with the observed mortality patterns across species, including measures of the degree of iso/anisohydry, photosynthesis, carbohydrate depletion, and hydraulic safety margins. Large losses of spring and summer whole-tree conductance (driven by belowground losses of conductance) and shallower rooting depths were associated with species that exhibited greater mortality. Based on this retrospective analysis, we suggest that species more vulnerable to drought were more likely to have succumbed to hydraulic failure belowground.

Interactions of predominant insects and diseases with climate change in Douglas-fir forests of western Oregon and Washington, U.S.A.

Forest disturbance regimes are beginning to show evidence of climate-mediated changes, such as increasing severity of droughts and insect outbreaks. We review the major insects and pathogens affecting the disturbance regime for coastal Douglas-fir forests in western Oregon and Washington State, USA, and ask how future climate changes may influence their role in disturbance ecology. Although the physiological constraints of light, temperature, and moisture largely control tree growth, episodic and chronic disturbances interacting with biological factors have substantial impacts on the structure and functioning of forest ecosystems in this region. Understanding insect and disease interactions is critical to predicting forest response to climate change and the consequences for ecosystem services, such as timber, clean water, fish and wildlife. We focused on future predictions for warmer wetter winters, hotter drier summers, and elevated atmospheric CO2 to hypothesize the response of Douglas-fir forests to the major insects and diseases influencing this forest type: Douglas-fir beetle, Swiss needle cast, black stain root disease, and laminated root rot. We hypothesize that 1) Douglas-fir beetle and black stain root disease could become more prevalent with increasing, fire, temperature stress, and moisture stress, 2) future impacts of Swiss needle cast are difficult to predict due to uncertainties in May-July leaf wetness, but warmer winters could contribute to intensification at higher elevations, and 3) laminated root rot will be influenced primarily by forest management, rather than climatic change. Furthermore, these biotic disturbance agents interact in complex ways that are poorly understood. Consequently, to inform management decisions, insect and disease influences on disturbance regimes must be characterized specifically by forest type and region in order to accurately capture these interactions in light of future climate-mediated changes.

Stomatal kinetics and photosynthetic gas exchange along a continuum of isohydric to anisohydric regulation of plant water status.

Species' differences in the stringency of stomatal control of plant water potential represent a continuum of isohydric to anisohydric behaviours. However, little is known about how quasi-steady-state stomatal regulation of water potential may relate to dynamic behaviour of stomata and photosynthetic gas exchange in species operating at different positions along this continuum. Here, we evaluated kinetics of light-induced stomatal opening, activation of photosynthesis and features of quasi-steady-state photosynthetic gas exchange in 10 woody species selected to represent different degrees of anisohydry. Based on a previously developed proxy for the degree of anisohydry, species' leaf water potentials at turgor loss, we found consistent trends in photosynthetic gas exchange traits across a spectrum of isohydry to anisohydry. More anisohydric species had faster kinetics of stomatal opening and activation of photosynthesis, and these kinetics were closely coordinated within species. Quasi-steady-state stomatal conductance and measures of photosynthetic capacity and performance were also greater in more anisohydric species. Intrinsic water-use efficiency estimated from leaf gas exchange and stable carbon isotope ratios was lowest in the most anisohydric species. In comparisons between gas exchange traits, species rankings were highly consistent, leading to species-independent scaling relationships over the range of isohydry to anisohydry observed.

Thermotolerance and heat stress responses of Douglas-fir and ponderosa pine seedling populations from contrasting climates.

Temperature and the frequency and intensity of heat waves are predicted to increase throughout the 21st century. Germinant seedlings are expected to be particularly vulnerable to heat stress because they are in the boundary layer close to the soil surface where intense heating occurs in open habitats. We quantified leaf thermotolerance and whole-plant physiological responses to heat stress in first-year germinant seedlings in two populations each of Pinus ponderosa P. and C. Lawson (PIPO) and Pseudotsuga menziesii (Mirb.) Franco (PSME) from climates with contrasting precipitation and temperature regimes. Thermotolerance of detached needles was evaluated using chlorophyll fluorescence (FV/FM, FO) and electrolyte leakage. PSME was more heat tolerant than PIPO according to both independent assessments of thermotolerance. Following exposure of whole seedlings to a simulated heat wave at 45 °C for 1 h in a growth chamber, we monitored FV/FM, photosynthesis, stomatal conductance, non-structural carbohydrates (NSCs) and carbon isotope ratios (δ13C) for 14 days. Heat treatment induced significant reductions in FV/FM in both species and a transient reduction in photosynthetic gas exchange only in PIPO 1 day after treatment. Heat treatment induced an increase in glucose + fructose concurrent with a decrease in starch in both species, whereas total NSC and sucrose were not affected by heat treatment. The negative relationship between glucose + fructose and starch observed in treated plants may be due to the conversion of starch to glucose + fructose to aid recovery from heat-induced damage. Populations from drier sites displayed greater δ13C values than those from wetter sites, consistent with higher intrinsic water-use efficiency and drought resistance of populations from drier climates. Thermotolerance and heat stress responses appeared to be phenotypically plastic and representative of the environment in which plants were grown, whereas intrinsic water-use efficiency appeared to reflect ecotypic differentiation and the climate of origin.

True Randomness from Big Data.

Generating random bits is a difficult task, which is important for physical systems simulation, cryptography, and many applications that rely on high-quality random bits. Our contribution is to show how to generate provably random bits from uncertain events whose outcomes are routinely recorded in the form of massive data sets. These include scientific data sets, such as in astronomics, genomics, as well as data produced by individuals, such as internet search logs, sensor networks, and social network feeds. We view the generation of such data as the sampling process from a big source, which is a random variable of size at least a few gigabytes. Our view initiates the study of big sources in the randomness extraction literature. Previous approaches for big sources rely on statistical assumptions about the samples. We introduce a general method that provably extracts almost-uniform random bits from big sources and extensively validate it empirically on real data sets. The experimental findings indicate that our method is efficient enough to handle large enough sources, while previous extractor constructions are not efficient enough to be practical. Quality-wise, our method at least matches quantum randomness expanders and classical world empirical extractors as measured by standardized tests.

Mapping 'hydroscapes' along the iso- to anisohydric continuum of stomatal regulation of plant water status.

The concept of iso- vs. anisohydry has been used to describe the stringency of stomatal regulation of plant water potential (ψ). However, metrics that accurately and consistently quantify species' operating ranges along a continuum of iso- to anisohydry have been elusive. Additionally, most approaches to quantifying iso/anisohydry require labour-intensive measurements during prolonged drought. We evaluated new and previously developed metrics of stringency of stomatal regulation of ψ during soil drying in eight woody species and determined whether easily-determined leaf pressure-volume traits could serve as proxies for their degree of iso- vs. anisohydry. Two metrics of stringency of stomatal control of ψ, (1) a 'hydroscape' incorporating the landscape of ψ over which stomata control ψ, and (2) the slope of the daily range of ψ as pre-dawn ψ declined, were strongly correlated with each other and with the leaf osmotic potential at full and zero turgor derived from pressure-volume curves.

Reliance on shallow soil water in a mixed-hardwood forest in central Pennsylvania.

We investigated depth of water uptake of trees on shale-derived soils in order to assess the importance of roots over a meter deep as a driver of water use in a central Pennsylvania catchment. This information is not only needed to improve basic understanding of water use in these forests but also to improve descriptions of root function at depth in hydrologic process models. The study took place at the Susquehanna Shale Hills Critical Zone Observatory in central Pennsylvania. We asked two main questions: (i) Do trees in a mixed-hardwood, humid temperate forest in a central Pennsylvania catchment rely on deep roots for water during dry portions of the growing season? (ii) What is the role of tree genus, size, soil depth and hillslope position on the depth of water extraction by trees? Based on multiple lines of evidence, including stable isotope natural abundance, sap flux and soil moisture depletion patterns with depth, the majority of water uptake during the dry part of the growing season occurred, on average, at less than ∼60 cm soil depth throughout the catchment. While there were some trends in depth of water uptake related to genus, tree size and soil depth, water uptake was more uniformly shallow than we expected. Our results suggest that these types of forests may rely considerably on water sources that are quite shallow, even in the drier parts of the growing season.