Consumer pressure, seed versus safe-site limitation, and plant population dynamics

Plants often suffer reductions in fecundity due to insect herbivory. Whether this loss of seeds has population-level consequences is much debated and often unknown. For many plants, particularly those with long-lived seedbanks, it is frequently asserted that herbivores have minimal impacts on plant abundance because safe-site availability rather than absolute seed number determines the magnitude of future plant recruitment and hence population abundance. However, empirical tests of this assertion are generally lacking and the interplay between herbivory, spatio-temporal variability in seed- or safe-site-limited recruitment, and seedbank dynamics is likely to be complex. Here we use a stochastic simulation model to explore how changes in the spatial and temporal frequency of seed-limited recruitment, the strength of density-dependent seedling survival, and longevity of seeds in the soil influence the population response to herbivory. Model output reveals several surprising results. First, given a seedbank, herbivores can have substantial effects on mean population abundance even if recruitment is primarily safe-site-limited in either time or space. Second, increasing seedbank longevity increases the population effects of herbivory, because annual reductions in seed input due to herbivory are accumulated in the seedbank. Third, population impacts of herbivory are robust even in the face of moderately strong density-dependent seedling mortality. These results imply that the conditions under which herbivores influence plant population dynamics may be more widespread than heretofore expected. Experiments are now needed to test these predictions. Received: 3 November 1999 / Accepted: 15 February 2000     

Effect of soil water stress on soil respiration and its temperature sensitivity in an 18-year-old temperate Douglas-fir stand

We analyzed 17 months (August 2005 to December 2006) of continuous measurements of soil CO₂ efflux or soil respiration (R_S) in an 18‐year‐old west‐coast temperate Douglas‐fir stand that experienced somewhat greater than normal summertime water deficit. For soil water content at the 4 cm depth (θ) > 0.11 m³ m^?3 (corresponding to a soil water matric potential of ?2 MPa), R_S was positively correlated to soil temperature at the 2 cm depth (T_S). Below this value of θ, however, R_S was largely decoupled from T_S, and evapotranspiration, ecosystem respiration and gross primary productivity (GPP) began to decrease, dropping to about half of their maximum values when θ reached 0.07 m³ m^?3. Soil water deficit substantially reduced R_S sensitivity to temperature resulting in a Q₁₀ significantly < 2. The absolute temperature sensitivity of R_S (i.e. dR_S/dT_S) increased with θ up to 0.15 m³ m^?3, above which it slowly declined. The value of dR_S/dT_S was nearly 0 for θ < 0.08 m³ m^?3, thereby confirming that R_S was largely unaffected by temperature under soil water stress conditions. Despite the possible effects of seasonality of photosynthesis, root activity and litterfall on R_S, the observed decrease in its temperature sensitivity at low θ was consistent with the reduction in substrate availability due to a decrease in (a) microbial mobility, and diffusion of substrates and extracellular enzymes, and (b) the fraction of substrate that can react at high T_S, which is associated with low θ. We found that an exponential (van't Hoff type) model with Q₁₀ and R₁₀ dependent on only θ explained 92% of the variance in half‐hourly values of R_S, including the period with soil water stress conditions. We hypothesize that relating Q₁₀ and R₁₀ to θ not only accounted for the effects of T_S on R_S and its temperature sensitivity but also accounted for the seasonality of biotic (photosynthesis, root activity, and litterfall) and abiotic (soil moisture and temperature) controls and their interactions.     

The Effects of Misspecification of a Regression Model

The consequences of the misspecification of a regression model are considered. For small effects of covariates a proportional consistency theorem is derived. The consistent estimation of the covariance matrix of the estimates is discussed.     

A Generalization of the GMANOVA-Model

It is shown that in some experimental designs the MANOVA- and the GMANOVA-model are too restrictive either to yield all hypothesis tests of interest or to reflect all known features of the design. An extension of these models is derived by relating the response vectors with the unknown model parameters by linear equations which may be completely different for each of the p components of the response vector and for each of the n independent vectors. For situations, in which a Wishart-distributed estimator for the underlying common covariance matrix is attainable, a test for any s-dimensional linear hypothesis on the model parameters is derived.     

Modeling productivity of a salt marsh

A model of primary productivity in a salt marsh is developed and compared to a regression analysis study of data showing dependence of growth on growing season, mean tidal height, and average monthly temperatures for several grass species.     

Testing against Ordered Alternatives in One-way Layout

In this paper we investigate a simple method of modifying well known nonparametric tests for the several samples location problem to yield a class of test statistics suitable for ordered alternatives. Optimum member of the class is identified in each case and its efficacy obtained. The method is applied to five statistics.     

Optimal Split-plot Designs for a Mixed Model

Exact tests are given, for the usual hypotheses on split-plot models with random blocks and fixed treatment effects, considering different numbers of blocks for each level of whole-plot treatment and assuming normally distributed observations. U- and D-optimal designs are considered with respect to the tests of main effects and interactions as well as to estimation of parameters.