The rise in spring temperatures over the past half‐century has led to advances in the phenology of many nontropical plants and animals. As species and populations differ in their phenological responses to temperature, an increase in temperatures has the potential to alter timing‐dependent species interactions. One species‐interaction that may be affected is the competition for light in deciduous forests, where early vernal species have a narrow window of opportunity for growth before late spring species cast shade. Here we consider the Marsham phenology time series of first leafing dates of thirteen tree species and flowering dates of one ground flora species, which spans two centuries. The exceptional length of this time series permits a rare comparison of the statistical support for parameter‐rich regression and mechanistic thermal sensitivity phenology models. While mechanistic models perform best in the majority of cases, both they and the regression models provide remarkably consistent insights into the relative sensitivity of each species to forcing and chilling effects. All species are sensitive to spring forcing, but we also find that vernal and northern European species are responsive to cold temperatures in the previous autumn. Whether this sensitivity reflects a chilling requirement or a delaying of dormancy remains to be tested. We then apply the models to projected future temperature data under a fossil fuel intensive emissions scenario and predict that while some species will advance substantially others will advance by less and may even be delayed due to a rise in autumn and winter temperatures. Considering the projected responses of all fourteen species, we anticipate a change in the order of spring events, which may lead to changes in competitive advantage for light with potential implications for the composition of temperate forests.

No abstract

Stepwise regression is often used to draw associations between phenological records and weather data. For example, the dates that a species first flowers each year might be regressed on monthly mean temperatures for a period preceding flowering. The months that 'best' explain the variation in first flowering dates would be selected by stepwise regression. However, daily records of weather are usually available. Stepwise regression on daily temperatures would not be appropriate because of high correlations between neighbouring days. Smoothing methods provide a way of avoiding such difficulties. Regression coefficients can be smoothed by penalising differences in slopes between neighbouring regressors. The resultant curve of regression gradients is intuitively attractive. Various possible approaches to smoothing regression coefficients are discussed. We illustrate the use of one method, P-spline signal regression, which is particularly appropriate when there are many more regressors than observations. Smoothing can be applied to more than one set of regressors. This results in a multi-dimensional surface of regression coefficients. We use this approach to investigate how the time of year that a plant species tends to flower affects its relationship with temperature records. Using this method, we found that later species tend to be affected by later temperatures.

A quantitative real-time PCR assay using TaqMan chemistry has been developed to quantify the level of Tilletia spp. contamination in wheat-seed lots. In the UK wheat seed is predominantly contaminated with Tilletia caries (syn. Tilletia tritici ), and the probability of detecting other Tilletia spp. is negligible. DNA standards, prepared from T. caries spores, were calibrated using a set of 26 seed samples, with T. caries contamination levels ranging from 0 to 1000 spores per seed. The linear calibration model obtained by the regression of log 10 (number of spores per seed + 1) on mean log 10 DNA ( µ g) produced a coefficient of determination ( R 2 ) of 0·904. The calibration model was tested using 226 seed samples; of these, 91% fell within the 95% confidence intervals. Of the 21 samples that were outside the limits, 16 were overpredictions and five underpredictions. The five underpredictions were all from seed samples where contamination was less than one spore per seed. The model predicts that samples with 44 pg of DNA will be below one spore per seed with 95% probability. Of the 226 test samples compared with this threshold, 99 contained less than 44 pg DNA, and these were found to have less than one spore per seed by microscopic assay. This real-time assay allows an increase in test throughput and provides the sensitivity required for an advisory threshold of one spore per seed.

No abstract