Multispecies deep learning using citizen science data produces more informative plant community models.

In the age of big data, scientific progress is fundamentally limited by our capacity to extract critical information. Here, we map fine-grained spatiotemporal distributions for thousands of species, using deep neural networks (DNNs) and ubiquitous citizen science data. Based on 6.7 M observations, we jointly model the distributions of 2477 plant species and species aggregates across Switzerland with an ensemble of DNNs built with different cost functions. We find that, compared to commonly-used approaches, multispecies DNNs predict species distributions and especially community composition more accurately. Moreover, their design allows investigation of understudied aspects of ecology. Including seasonal variations of observation probability explicitly allows approximating flowering phenology; reweighting predictions to mirror cover-abundance allows mapping potentially canopy-dominant tree species nationwide; and projecting DNNs into the future allows assessing how distributions, phenology, and dominance may change. Given their skill and their versatility, multispecies DNNs can refine our understanding of the distribution of plants and well-sampled taxa in general.

AFLP markers reveal high clonal diversity and extreme longevity in four key arctic-alpine species.

We investigated clonal diversity, genet size structure and genet longevity in populations of four arctic-alpine plants (Carex curvula, Dryas octopetala, Salix herbacea and Vaccinium uliginosum) to evaluate their persistence under past climatic oscillations and their potential resistance to future climate change. The size and number of genets were determined by an analysis of amplified fragment length polymorphisms and a standardized sampling design in several European arctic-alpine populations, where these species are dominant in the vegetation. Genet age was estimated by dividing the size by the annual horizontal size increment from in situ growth measurements. Clonal diversity was generally high but differed among species, and the frequency distribution of genet size was strongly left-skewed. The largest C. curvula genet had an estimated minimum age of c. 4100 years and a maximum age of c. 5000 years, although 84.8% of the genets in this species were <200 years old. The oldest genets of D. octopetala, S. herbacea and V. uliginosum were found to be at least 500, 450 and 1400 years old, respectively. These results indicate that individuals in the studied populations have survived pronounced climatic oscillations, including the Little Ice Age and the postindustrial warming. The presence of genets in all size classes and the dominance of presumably young individuals suggest repeated recruitment over time, a precondition for adaptation to changing environmental conditions. Together, persistence and continuous genet turnover may ensure maximum ecosystem resilience. Thus, our results indicate that long-lived clonal plants in arctic-alpine ecosystems can persist, despite considerable climatic change.

Longevity of clonal plants: why it matters and how to measure it.

BACKGROUND: Species' life-history and population dynamics are strongly shaped by the longevity of individuals, but life span is one of the least accessible demographic traits, particularly in clonal plants. Continuous vegetative reproduction of genets enables persistence despite low or no sexual reproduction, affecting genet turnover rates and population stability. Therefore, the longevity of clonal plants is of considerable biological interest, but remains relatively poorly known. SCOPE: Here, we critically review the present knowledge on the longevity of clonal plants and discuss its importance for population persistence. Direct life-span measurements such as growth-ring analysis in woody plants are relatively easy to take, although, for many clonal plants, these methods are not adequate due to the variable growth pattern of ramets and difficult genet identification. Recently, indirect methods have been introduced in which genet size and annual shoot increments are used to estimate genet age. These methods, often based on molecular techniques, allow the investigation of genet size and age structure of whole populations, a crucial issue for understanding their viability and persistence. However, indirect estimates of clonal longevity are impeded because the process of ageing in clonal plants is still poorly understood and because their size and age are not always well correlated. Alternative estimators for genet life span such as somatic mutations have recently been suggested. CONCLUSIONS: Empirical knowledge on the longevity of clonal species has increased considerably in the last few years. Maximum age estimates are an indicator of population persistence, but are not sufficient to evaluate turnover rates and the ability of long-lived clonal plants to enhance community stability and ecosystem resilience. In order to understand the dynamics of populations it will be necessary to measure genet size and age structure, not only life spans of single individuals, and to use such data for modelling of genet dynamics.