Beyond white noise: the colors of noise in long-term COM(P)ADRE data

When ecologists build simulations of populations of plants or animals, we usually add in randomness, because populations are always changing in ways that are impossible to fully pre-determine. This randomness is usually white noise, which means that the random values at each point in time are uninformed by the values that came before. Each time, the slate is wiped clean and a new value is drawn from scratch. This is a nice and easy way to generate randomness, but the problem is that populations usually don’t fluctuate with white noise, but with red noise.

Screenshot 2019-11-09 at 13.31.37

Red noise is a kind of randomness where each random value is positively correlated with the ones that came before. That is to say, it’s randomness with a memory: when it starts to go on a downward trend, it tends to continue that downward trend. You can see this in the images above, where the red noise is much less “wiggly” than the white noise. Blue noise is the opposite of red noise: it also has a memory, but it is negatively correlated to previous points and tends toward the opposite of what happened before.

Population ecologists have already studied colored noise in population growth. Populations usually exhibit red noise in their fluctuations. This is important to know because red noise tends to increase extinction risk: if a population size happens to wiggle downward, then it tends to continue in that direction, until it might eventually crash. But overall population size is just one part of the picture. Population size comes from the combination of births (fertility) and deaths (mortality) of individuals of different ages and life stages. In our study, we wanted to find out if fertility and mortality in different life stages exhibit colored noise in their year-to-year fluctuations, and if they do, how that might change the way we simulate populations.

Nature is full of colored noise, but it can be hard to measure, because you need at least fifteen continuous time points to get an accurate read on noise color, and most datasets on populations are short-term. Fortunately, the COMADRE and COMPADRE databases have detailed population data on a variety of organisms, some of them long-term. We gathered all the long-term datasets we could find, and measured the noise color in fertility and mortality at different life stages. Here are two examples.

Screenshot 2019-11-09 at 13.31.42

The one on the left is the mountain gorilla, and the one on the right is a plant called velvety goldenrod. Each column in the matrix represents an age, youngest at the left and oldest at the right (velvety goldenrod usually only lives one or two years.) The first row of the matrices is fertility at that age, and the other rows are survival. In both of these species, you can see that fertility and survival have very different noise colors.

Each matrix is also outlined in a color that indicates the noise color for overall population growth. You can see that the noise color for the population growth isn’t always obvious from the noise color of individual fertility and survival squares within the matrix. In the gorilla, the population growth is strongly red noise, like the fertility rates, while in the goldenrod the growth rate has nearly white noise, somewhere in between the blue noise of the fertility and the red noise of survival.

We hope our work will open the door for other researchers to learn more about colored noise in fertility and survival. To that end, we created a package in R for modeling populations with colored noise. It’s called, appropriately, colorednoise.

Julia Pilowsky

The findings and applications of this work can be found in a recent publication here:

Pilowsky & Dahlgren. 2019. Incorporating the temporal autocorrelation of demographic rates into structured population models. Oikos DOI 10.111/oik.06438


Can demographic data be borrowed across species? Probably not for plants

A common challenge in conserving endangered species is not knowing enough about the species itself. Information about the basic biology of the threatened species is critical for helping to determine how well the species is doing and exploring possible strategies for recovery. In particular, information on survival and fecundity rates at different life stages can be used to identify which stages may need the most protection. Demographic rates can also be used to project what the population will look like in the future and under different possible recovery strategies. Unfortunately, data on survival and fecundity rates are rarely available, especially for endangered species, because they are gathered through detailed population monitoring over several years or more. In absence of these data, managers and scientists often suggest “borrowing” data from similar species to fill those data gaps. For this to work, however, the species that are considered similar would need to have demographic rates that are nearly the same or follow similar patterns.

In our recently published study, we tested this assumption. We asked whether plant species that are similar (closely related and/or have similar biological traits) would have similar demographic rates. We used matrix population models from the COMPADRE Plant Matrix Database to estimate three types of demographic rates: intrinsic population growth, stage-specific survival, and fecundity rates, for populations from 425 plant species. We constructed a hierarchical Bayesian model to predict these demographic rates based on maximum plant height and as many life history traits as we were able to find (which was unfortunately not very many for plants!). We also estimated the strength of the phylogenetic signal in these rates, which would tell us the extent to which closely related species would have similar demographic rates. By modelling the different populations for the same species hierarchically, we were also able to look at how demographic rates varied within a species compared to among different species. This provides insight into whether it would be more appropriate to borrow data from a different population of the same species compared to borrowing data from a similar species.

We found that demographic rates were not well-predicted based on phylogenetic relatedness and traits. Phylogenetic signal was weak for population growth and survival rates, but was relatively strong for fecundity rates (Pagel’s lambda = 0.71, 95% CI = 0.43, 0.85). Demographic rates also varied greatly among populations of the same species. For intrinsic population growth, the largest sources of variation in our model were the unexplained variation among species (after accounting for traits and phylogeny), and unexplained variation among populations of the same species (compare “species error” and “population error” in Fig. a; taken from Fig. 2 in the paper). This indicates the traits we included in the model were not good predictors of population growth, and adding other species-level traits may improve the predictions for this demographic rate. However, intrinsic growth rates also varied a lot among populations of the same species. For the other demographic rates we tested, the unexplained variation among populations was actually higher than that among species (e.g., for adult survival shown in Fig. d). This means that differences in survival and fecundity rates among populations of the same species was as great or greater than the differences among different species. This finding suggests that these vital rates are highly dependent on population-level factors such as the specific environmental conditions for the given location and year.

Screen Shot 2018-08-27 at 17.42.17.png

Our results suggest that borrowing data from similar species, or even a different population of the same species, may not be useful in a conservation context. This is so because the demographic rates for a closely related or similar species could actually be very different from that of the focal species, and using those rates could lead to very different conclusions about how a population is doing. These findings are consistent with previous studies that have asked similar questions, yet the idea of borrowing data continues to crop up because the lack of data is so pervasive. Demographic monitoring can be costly and time consuming, and therefore it is rarely done (or done well), especially for threatened species. We hope our study can help to promote more direct species monitoring to help inform species conservation and management.

Judy Che-Castaldo

Research Scientist at Lincoln Park Zoo – Chicago, USA

Core committee member of COMPADRE & COMADRE



Che-Castaldo, J.P., C.C. Che-Castaldo, M.C. Neel. 2018. Predictability of demographic rates based on phylogeny and biological similarity. Conservation Biology

Embracing sampling uncertainty in analyses with COM(P)ADRE

by Patrick Barks (University of Southern Denmark, email:

The COM(P)ADRE Plant and Animal Matrix Databases together contain thousands of population projection matrices from hundreds of individual studies. The availability of these matrices to researchers has led to fascinating comparative analyses in the fields of ecology, evolution, and demography, at taxonomic, spatial, and temporal scales that would not otherwise be possible (see here for a list of relevant publications).

One of the challenges inherent in such analyses is that it’s often difficult to obtain information regarding the degree of sampling uncertainty associated with the values that populate projection matrices (i.e. stage- or age-specific transition rates based on survival, growth, and reproduction). These transition rates are almost always estimates of population parameters based on samples (population in the statistical sense), and therefore have associated sampling uncertainty, as do any parameters derived from them (e.g. population growth rate, damping ratio, life expectancy, etc.)1. Transition estimates based on a small number of individuals will tend to have large uncertainty, while those based on larger samples have less uncertainty. For example, the figure below depicts the sampling uncertainty associated with a stage-specific survival rate of 40% estimated from a random sample of N = 5 individuals vs. N = 50 individuals.

sampling uncertainty

Whereas sampling uncertainty is routinely incorporated into statistical analyses in the original studies that produce projection matrices, it is rarely incorporated into analyses that use published projection matrix data from sources like COM(P)ADRE. This omission may lead to bias or overconfidence in some types of analyses.

To investigate this possibility, we are initiating a study to examine the nature and distribution of sampling uncertainty among projection matrices in the COMPADRE Plant Matrix Database. Our goals are to:

  • understand whether uncertainty in transition rates is relevant for analyses based on COMPADRE,
  • assess the types of variables or analyses that are most likely to be affected by sampling uncertainty, and
  • develop resources to help researchers incorporate sampling uncertainty into their analyses.

To this end, we are currently working to obtain additional data for as many of the matrices in COMPADRE as possible, with a specific focus on matrices from unmanipulated, wild populations. To estimate sampling uncertainty we generally require more information than is available in the original papers (e.g. stage-specific sample sizes, counts of reproductive structures, etc.), so we will be contacting many authors over the coming weeks to request these data. We sincerely appreciate the time and effort taken by researchers to make their hard-won datasets available to us. The inclusion of these matrices in COMPADRE is already a great service to the scientific community and we hope that our study will further increase their utility to researchers, and help to improve inferences derived from the COM(P)ADRE databases.

An example analysis using COMPADRE

To make the issue of sampling uncertainty more concrete, we’ll work through an example analysis with COMPADRE. Specifically, we’ll test the hypothesis that relatively long-lived species tend to experience relatively low year-to-year variation in population growth rates (λ). For simplicity, we’ll limit this analysis to species categorized as herbaceous perennials, and unmanipulated populations with at least three annual transition matrices in COMPADRE (and a few more selection criteria noted in the RMarkdown document here).

variance in lambda as function of life expectancy

In the figure above, each point represents a population (as defined in COMPADRE), and the best-fit line is from a linear mixed model that accounts for non-independence of populations from the same species. There are of course different modeling approaches we could have taken — estimate life expectancy at the species level rather than population level, use a more complete model of phylogenetic non-independence, etc. — but we’ll save some of that for later.

For now we’d like to know, how wide are the error bars associated with each point in the figure above? The regression model assumed zero uncertainty in both life expectancy and variance(log λ), but as previously noted, both variables are estimates of population parameters with inherent sampling uncertainty. Let’s take a detour here to try to estimate sampling uncertainty for a single population.

Modeling uncertainty in transition rates

Consider a set of matrices available in COMPADRE from a 6-year study of the perennial forb Agrimonia eupatoria (Rosaceae) in southern Sweden (Kiviniemi 2003)2. The matrices give us point estimates for each transition rate in each year, which we can use to calculate point estimates for derived parameters such as life expectancy, λ, and variance(log λ). But to estimate the uncertainty in all these parameters, we need information from outside COMPADRE3.

First, we need to know how the transition rates were estimated. Based on the original paper, survival transitions were estimated directly from the fates of marked individuals (i.e. Aij = number transitioned from stage i in year t to stage j in year t+1 / number in stage i in year t), and the single fecundity transition was estimated using the anonymous reproduction method (i.e. Asr = number in seedling stage in year t+1 / number in reproductive stage in year t). Given this methodology, to reconstruct the raw counts from which each transition rate was estimated, all we need are the denominators in the equations above (i.e. stage-specific sample sizes for each transition period), which the original paper helpfully provides.

The figure below shows point estimates for each transition rate (open circles), as well as 90% and 99% confidence intervals based on the relevant sampling distribution (thin and thick bars; assuming multinomial and Poisson distributions for the survival and fecundity transitions, respectively)4.

Estimates and CI for each transition rate

To estimate the sampling distributions of the derived parameters, we generate thousands of simulated projection matrices by repeatedly drawing from the sampling distribution of each transition rate. The sampling distributions for the derived parameters are summarized below, again alongside the corresponding point estimates.

Sampling distributions for the derived paramet

For some transition periods, the point estimate for life expectancy is quite far from the respective confidence interval. This can occur when the sample of individuals in one or more stage classes experiences 100% survival, in which case the point estimate for life expectancy may be very high. But the sampling distribution for those one or more survival parameters will only include values ≤ the point estimate (e.g. see the 1996-97 seedling-to-juvenile transition)5.

Incorporating sampling uncertainty into our analysis

We can now add the sampling uncertainty for the population of Agrimonia eupatoria to our original figure.

Sampling uncertainty for the population of Agrimonia eupatoria

On one hand, the sampling uncertainty for this population seems high. On the other hand, we have a lot of data, and the relationship between life expectancy and variance(log λ) is strong. If we extrapolate (wildly) from this and a few other populations for which we have data, to make simple assumptions about the distribution of sampling uncertainty among all populations, we’ll find that the observed degree of uncertainty is unlikely to change the results of the current analysis. The figure below depicts predictions from an extension of the previously-described mixed effect model that now also incorporates simulated measurement error in both life expectancy and variance(log λ). The results are essentially unchanged.

Figure incorporating simulated measurement error

Perhaps this will be the case for many analyses with COMPADRE. But perhaps some types of analyses based on smaller subsets of data, or with more marginal effect sizes, will be more strongly influenced by sampling uncertainty. Either way, we think it warrants investigation, and we hope to report back with the answer.


Kiviniemi, K. (2002). Population dynamics of Agrimonia eupatoria and Geum rivale, two perennial grassland species. Plant Ecology, 159, 153-169.


1Some of the matrices in COM(P)ADRE may in fact be based on data from entire biological populations rather than samples. Whether these map to ‘statistical populations’ will depend on the research question.

2Kiviniemi (2003) studied two populations of A. eupatoria, denoted A and B. Our analysis only includes population B, because the annual matrices for population A were mostly non-ergodic.

3Apart from uncertainty in the underlying transition rates, the uncertainty in variance(log λ) is also a function of the number of years over which it was estimated. This latter component of uncertainty is straightforward to model, but we ignore it here for simplicity.

4Because the transition rates reported in the original paper were estimated independently across years and transition types, our estimates of sampling error make the same assumptions. But now that we’ve reconstructed the raw data, we could of course model the transitions using a more nuanced correlational structure — e.g. partially pooling across years, or allowing for correlations among transition types.

5Note also that the point estimate for life expectancy for the 1994-95 transition was incalculable, because the estimated transition rates implied a 100% survival loop between the final two stage classes (i.e. infinite life expectancy).

Demography beyond a dissertation: profiting from well replicated and long term data

Although there are thousands of studies of population dynamics (Crone et al. 2011, Merow et al. 2014), most occur at small spatial scales and span few years (Menges 2000, Crone et al. 2011, Salguero-Gómez et al. 2015). We are persuaded that limited spatial replication and short study intervals can hinder our ability to adequately understand and predict population dynamics. In a recent publication (Quintana-Ascencio et al. accepted) we used an unusually detailed, spatially expansive, long-term dataset to unravel complex interactions between landscape patterns and ecological disturbances affecting a species’ distribution and demography. We assessed the effects of landscape factors on the population dynamics of Hypericum cumulicola, a pyrogenic, endangered Florida scrub endemic plant species. We took advantage of an ongoing 22-year study, that has monitored many populations (sampled in 15 independent habitat patches) with a rich disturbance history (14 independent fires), well-described habitat requirements (for open gaps in Florida scrub dominated by Ceratiola ericoides) and a strong spatially patterned landscape.

Screen Shot 2018-04-05 at 12.21.13

We use of a long-term dataset to decouple effects of disturbance from population and year effects. The effects of disturbance regimes on population dynamics can be challenging to study because of the large amount of longitudinal data required. Long-term studies may be necessary to avoid misleading conclusions built on chronosequence-based short-term studies. Even when such long-term data are available, decoupling the effects of time since disturbance from year and population effects (e.g. climate, biotic interactions) requires datasets with replication across space and time.

We have shown before that H. cumulicola vital rates were strongly related to fire, the predominant ecological disturbance. Yet, in the reference study, we provide further evidence that vital rates were also affected by small scale landscape patterns such as elevation that alter the distance to the water table and larger scale patterns of patches in the landscape such as patch aggregation (e.g. survival was higher in large aggregated patches).

Screen Shot 2018-04-05 at 12.22.27

Overall, our results indicated the importance of metapopulation dynamics. H. cumulicola occupancy and population growth predictions were highest in large, aggregated habitat patches. Large habitat patches may have several advantages for metapopulation dynamics, including a larger target for rescues, greater diversity of microhabitats, and the ability to support larger populations less prone to demographic or environmental stochasticity (Hanski, Moilanen & Gyllenberg 1996). Larger habitat patches may also be more likely to support larger populations of the dominant shrub Florida rosemary, which may act as a foundation species for H. cumulicola and other herbaceous plants that are vulnerable to competition from resprouting shrubs. Model predictions of abundance were most reliable for patches with the extremes of patch area or aggregation. In this study, predictions of occupancy and abundance were frailer for habitat patches with small and intermediate size and aggregation. We have previously documented that unoccupied patches can be suitable habitat for H. cumulicola. These results indicate that limited dispersal and unfavorable matrix habitats can synergistically contribute to colonization failure. Few studies have integrated population models and landscape level environmental drivers to characterize species distributions, although this approach has great promise for assessing the consequences of environmental changes

The distribution of species across landscapes ultimately reflects the interaction of demography with landscape and disturbance properties. However, both demographic inertia (e.g., long life span, dormant stages) and landscape history (e.g. environmental legacies) may create lags in responses. Therefore, realized species distributions (reflecting past interactions) may have different patterns than current vital rates. We used this detailed long-term monitoring to assess the relationships between demography and landscape-level drivers to disentangle the potential causes of realized species distributions. H. cumulicola occupancy peaked at higher elevations in larger patches but many vital rates peaked at lower elevations. This may reflect lags in demography such as the role of seed banks in allowing populations to persist between disturbances. Better understanding of the spatial and temporal dynamics of seed dormancy and dispersal and the role of environmental factors on their variation will greatly benefit our understanding of regional plant population persistence. In addition, landscape changes in extreme microsites (e.g. the largest open patches) may lag behind other landscape patterns. Short-term or limited studies in areas with these persisting patches may be biased, as patches where all plants have died cannot be a source of data. This demographic ghost of mortality past may explain unexpected demographic patterns in chronosequence studies.

Screen Shot 2018-04-05 at 12.23.17

Many critiques of demographic modeling have pointed out that models do a poor job of predicting beyond more than a few years (Crone 2013). By incorporating common drivers such as weather, habitat suitability, disturbances, landscape patterns, and land management, we can improve our understanding of temporal and spatial variability in demography (Ehrlén & Morris 2015). Weaknesses of population models can be attributed to their frequent dependence on short spans of data (Menges 2000) and on their frequent blindness to underlying drivers (Ehrlén & Morris 2015). The scaling up of local demography to landscapes, accomplished by a combination of approaches, can lead to more effective management for ecological diversity.

Pedro F Quintana Ascencio and Eric S Menges




Crone, E.E., et al. (2011) How do plant ecologist use matrix population models? Ecology Letters, 14, 1-8.

Ehrlén, J. & Morris, W.F. (2015) Prediction changes in the distribution and abundance of species under environmental change. Ecology Letters, 18, 303-314.

Hanski, I., Moilanen, A. & Gyllenberg, M. (1996) Minimum viable metapopulation size. American Naturalist, 147, 527-541.

Menges, E.S. (2000) Population viability analyses in plants: challenges and opportunities. Trends in Ecology and Evolution 15, 51-56.

Merow, C. et al. (2014) Advancing population ecology with integral projection models: a practical guide. Methods in Ecology and Evolution 5, 99-110.

Quintana-Ascencio, P.F., Koontz, S., Smith, S. Sclater, V. David, A., Menges, E. S. (2018) Predicting landscape-level distribution and abundance: Integrating demography, fire, elevation, and landscape habitat configuration. Journal of Ecology, Accepted manuscript online: 2 APR 2018 12:00AM EST | DOI: 10.1111/1365-2745.12985

Salguero-Gómez, et al. (2015) The COMPADRE plant matrix database: an open online repository for plant demography. Journal of Ecology 103, 202-218.








Attacking the question of vegetative dormancy with COMPADRE

Vegetative dormancy is the tendency that some herbaceous plants have to forego sprouting for one or more years at a stretch. It is a remarkably common condition among the terrestrial plants, having been found in over 20 plant families, and over 100 plant species. The real number is likely to be far higher, because our knowledge of its extent is limited by the availability of demographic studies documenting it. The mechanisms responsible are also not fully understood, and likely vary across plant species.

The first studies to document dormancy that I am aware of go back at least to the 1940s, with work by Carl Olaf Tamm (Tamm, 1948). Carl was developing demographic studies of many European perennials and often noted that some individuals seemed to “disappear” for varying numbers of years from the aboveground population. In some cases, these absences were explained in terms of high levels of herbivory, while in others, no explanation was evident, except that the plants just did not seem to sprout. Intriguingly, these absences were most often noted in orchids, as in Carl’s work and later work by others (Hutchings, 1987; Tamm, 1956; Wells, 1967, 1981). This pattern led John Harper to note in his famous plant ecology textbook: “An odd feature of the depletion curves for the orchids is that the number of survivors appears to go up as well as down! Clearly the number of survivors can never increase. The explanation is that the orchids appear to be capable of disappearing from the above-ground population for a year, or perhaps two. …It may be that this habit is more common than we know” (Harper, 1977).



The yellow lady’s slipper orchid (Cypripedium calceolus), which grows throughout Eurasia, is a strongly dormancy-prone species of conservation concern.

Visions of undead plants aside, this particular quote by Harper hints at the crux of the issue with vegetative dormancy: just as the dead cannot go back to life, the dormant also cannot photosynthesize and (sexually) reproduce. Yet, there is a tendency to think of plants as lovely, green things that produce pretty flowers… except of course for those plants that cause us allergies and other inconveniences. Further, without reproduction, isn’t dormancy a time when evolutionary fitness drops (i.e. no reproduction!)? Dormancy therefore tests our own common sense, both because it forces us to acknowledge that plants are not necessarily organisms that need to produce leaves and photosynthesize, and because it makes us consider how exactly a seemingly maladaptive condition could be so common.

Our recent paper in Ecology Letters is an attempt to understand dormancy in a way that might lead to answers (Shefferson et al., in press). Up until this study, studies of dormancy were typically demographic studies of one plant species, indeed typically just one population. Here, we developed our own large dataset of characteristics of these studies, and of the plant species and populations that were under investigation in each case. We strengthened this literature-based dataset with several tens of individual demographic datasets which we analyzed in a standardized way. These standardized analyses included generalized linear mixed model assessments of the impacts of size, reproduction, and sprouting on all vital rates estimable from each dataset, and covering not simply two year intervals (as is the case in virtually all plant demographic studies), but three, in order to better assess the impacts of individual history on vital rates. The three-year approach to analysis is one that I have worked with a great deal and find very useful, because it allows the exploration of the impacts of growth trajectories on vital rates, for example the impacts of large, rapid growth between two years on future survival or reproduction (Shefferson, Warren II, & Pulliam, 2014). We also estimated mean life expectancy, maximum (and median) dormancy duration, and the mean proportion of plants dormant per year, as well as a number of other metrics. This served as the raw material for an analysis of roughly 300 populations and over 100 plant species as to what sorts of patterns we could find in dormancy at the global level.

2012 May - Chicago 068.JPG

A volunteer steward monitors the rare orchids of a county forest preserve near Chicago, Illinois, USA.

Our results were striking, and most notably that dormancy does appear to be driven by common factors across the plant kingdom. Precipitation levels, temperature experienced, herbivory levels, and even latitude determined the levels of dormancy observed, as did the nutritional status of the plant, and its growth form. Trade-offs were strongly linked to dormancy, most notably whether there were any costs of sprouting, such as increased mortality due to greater visibility to herbivores, or a loss of future reproduction due to exhaustion of resources on a sprout in a given year. Costs of growth were also evident, and were documented as costs to survival, reproduction, or sprouting on the basis of growth (i.e. positive change in size) between the previous two years. These results should give biologists new leads to pursue in addressing the ultimate mechanisms behind dormancy.

The COMPADRE Plant Matrix Database was a big help in this analysis. Although the main analyses were performed using our own dataset, COMPADRE was used in two ways. First, in a number of cases, the authors of studies incorporated in our own study did not make their demographic data available publically except through COMPADRE. Thus, in about 20-30 cases, we used projection matrices documented in COMPADRE to supplement what we know about a particular study from the associated paper. Second, we used COMPADRE to test the hypothesis that vegetative dormancy has a common evolutionary origin. We did this by supplementing our own dataset with the identities of herbaceous perennial plant species noted in COMPADRE as not exhibiting dormancy. We then assessed the evolution of vegetative dormancy across the plant kingdom by mapping dormancy status onto a tree of all of these species, procured from the Open Tree of Life Project (Michonneau, Brown, & Winter, 2016). This analysis led us to discover that there were likely many evolutionary gains, as well as losses, of dormancy, and thus that there are likely many genetic bases to the trait.

In the end, our project did something that studies of individual populations could not do – find a common signal across wild-collected data in a typically unobservable life stage, showing what exactly drives this fundamentally interesting but very bizarre phenomenon. We believe this macroecological perspective can greatly alter our understanding of the natural world, and we welcome further studies along these lines.

Richard P. Shefferson

University of Tokyo



Harper, J. L. (1977). Population biology of plants. New York, New York, USA: Academic Press.

Hutchings, M. J. (1987). The population biology of the early spider orchid, Ophrys sphegodes Mill. I. A demographic study from 1975 to 1984. Journal of Ecology, 75(3), 711–727.

Michonneau, F., Brown, J. W., & Winter, D. J. (2016). rotl: an R package to interact with the Open Tree of Life data. Methods in Ecology and Evolution, 7(12), 1476–1481. doi:10.1111/2041-210X.12593

Shefferson, R. P., Kull, T., Hutchings, M. J., Selosse, M.-A., Jacquemyn, H., Kellett, K. M., … Whigham, D. F. (in press). Drivers of vegetative dormancy across herbaceous perennial plant species. Ecology Letters. doi:10.1111/ele.12940

Shefferson, R. P., Warren II, R. J., & Pulliam, H. R. (2014). Life history costs make perfect sprouting maladaptive in two herbaceous perennials. Journal of Ecology, 102(5), 1318–1328. doi:10.1111/1365-2745.12281

Tamm, C. O. (1948). Observations on reproduction and survival of some perennial herbs. Botaniska Notiser, 1948(3), 305–321.

Tamm, C. O. (1956). Further Observations on the Survival and Flowering of Some Perennial Herbs, I. Oikos, 7(2), 273–292. doi:10.2307/3564927

Wells, T. C. E. (1967). Changes in a population of Spiranthes spiralis (L.) Chevall. at Knocking Hoe National Nature Reserve, Bedfordshire, 1962-65. Journal of Ecology, 55, 83–99.

Wells, T. C. E. (1981). Population ecology of terrestrial orchids. In H. Synge (Ed.), The biological aspects of rare plant conservation (pp. 281–295). New York, New York, USA: John Wiley & Sons.

Introducing our brand new research network coordinator

Hola a todos! My name is Haydée Hernández Yáñez, and I am the new research coordinator for the COMPADRE/COMADRE database. I am very happy to have recently joined the team. I am keen to contribute to the expansion of this remarkable repository of matrix population matrices for plants and animals under the auspices of a recent NSF grant on biotechnology.


I am an ecologist interested in conservation science, spatial ecology, and species interactions. Through my research, I have worked with both plants and animals, in the field and on data analyses. While I was an undergraduate, I worked as a field assistant in the Yucatan Peninsula in Mexico, collecting blood samples from Yucatan wrens (Campylorhynchus yacatanicus), a bird species endemic to Mexico. I also worked as a research assistant collecting field information on plant interactions with ants, as well as constructing mutualistic networks of floral visitors and their host plants. I thoroughly enjoyed working on network analysis and learning about the importance of species interactions. Following my passion for plant-animal interactions, I enrolled in a MSc at the University of Missouri-St. Louis, where I researched the role of herbivores and soil on plant distribution and abundance in a Costa Rican wet forest. It was a great opportunity for me, as I learned more about the role of biotic and abiotic factors in shaping communities.

After graduating, I took a job as a research intern at the GIS lab of the Smithsonian Conservation Biology Institute. There, I applied statistical models to understand the geographical distribution and habitat requirements of the giant panda and Arabian bustards. I wholeheartedly enjoyed working in the area of spatial ecology and with animals (granted, I only saw their data points)! During my time there, I also worked on the database of the Smithsonian’s Movement of Life Initiative, which collects movement data received from collared animals in the field, such as Scimitar-Horned Oryx in Chad and Asian elephants in Myanmar. This experience has made me realize how important spatial ecology is as a tool for conservation science.

I’m currently catching up with the vast literature on population biology and matrix population models. It is a rather steep learning curve (e.g. see the lengthy data-entry protocol, and user guidelines of COMPADRE and COMADRE), but I am convinced that soon I’ll be fully caught up and able to actively shape COMPADRE and COMADRE into new exciting directions. I look forward to working with the team, data contributors, and users to make of the COM(P)ADRE databases an even better repository that will help us in our continual journey of understanding species’ biology.

Haydée Hernández Yáñez


Twitter: @Haydeshka


What makes some plant species more vulnerable to extinction than others? With the current extinction crisis in mind, this question is becoming increasingly important. Because for more than 90% of known plant species the current conservation status is still unassessed (IUCN, 2017) and with a lot of species yet to be discovered (Pimm & Raven, 2017), understanding the factors that are associated a high extinction vulnerability may help to prioritize populations for conservation. Examples of such factors are low reproductive rates and a high susceptibility to environmental disturbance. While there has been considerable effort in parameterizing and compiling demographic data, as demonstrated by the COMPADRE/COMADRE initiative, such data are available only for a relatively small amount of species.

This leads us to the start of our research (de Jonge et al., 2018), which was recently published in Biological Conservation. Inspired by the theory of allometric scaling (West, Brown, & Enquist, 1999), and by the recent work by one of our colleagues on mammals and birds (Hilbers et al., 2016), we decided to attempt to use functional traits to estimate demographic rates, and subsequently extinction risks, in plants. To do so, we derived the maximum growth rate for many species from the COMPADRE Plant Matrix Database (Salguero-Gómez et al., 2015), and  then we combined these with maximum plant height reported in the TRY database (Kattge et al., 2011). To estimate the susceptibility to environmental disturbances, we also calculated the variance in population growth rate for time-series reported in COMPADRE. As expected, plant height is negatively related to both maximum population growth rates and the susceptibility to environmental disturbances.

Screen Shot 2018-02-22 at 20.05.35

Figure 1. Relationships between plant height and maximum population growth rate and variance in population growth rate derived from combining the COMPADRE Plant Matrix Database with the TRY plant trait archive (de Jonge et al., 2018).

So now we know that large plant species indeed grow slower but also more stable than their smaller counterparts. But what does this mean for extinction vulnerabilities? Generally, being less susceptible to disturbances is better. However, being able to increase population sizes quickly can help populations recover after environmental disturbances. To see how these relationships with plant height add up, we calculated the Mean Time to Extinction (MTE) and Probability of Extinction (PE) as a function of height. And there you have it… bigger does seem to be better… all else being equal. Of course there are plenty of other variables that are probably more important than height, but these might not be measured as easily as height.

Screen Shot 2018-02-22 at 20.05.46

Figure 2. The probability of extinction (PE) as a function of plant height calculated using the derived relationships between demographic rates and plant height (de Jonge et al., 2018). Solid black line gives the median PE and shading shows the 90% confidence interval. Two plant species for which the PE could be calculated directly are shown, left: Solidago mollis and right: Eremophilia maitlandi

Our work demonstrates how linking large databases such as COMPADRE/COMADRE and functional trait databases (see also Adler et al. 2014) can reveal relationships between traits and demography. While COMPADRE already offers a huge collection of plant demographic, long-term monitoring studies on populations of large plants are still in short supply. Furthermore, during our research, we often found that the overlap in species between the different databases was too small to successfully combine them. For example, we were not able to investigate the role of other traits, such as seed mass or wood density, because there were not enough species occurring in COMPADRE for which such data were available in TRY. Increasing the species overlap might be done by integrating the demography and functional traits of species in one database. Such, multi-faceted global databases would help us to better understand the relationships between functional traits and demography.

Melinda M.J. de Jonge
PhD Student
Department of Environmental Science
Radboud University Nijmegen
PO Box 9010, 6500 GL Nijmegen, The Netherlands

Adler, P. B., Salguero-Gómez, R., Compagnoni, A., Hsu, J. S., Ray-Mukherjee, J., Mbeau-Ache, C., & Franco, M. (2014). Functional traits explain variation in plant life history strategies. Proc. Nat. Acad. Sci. USA, 111(2), 740-745. doi: 10.1073/pnas.1315179111

de Jonge, M. M. J., Hilbers, J. P., Jongejans, E., Ozinga, W. A., Hendriks, A. J., & Huijbregts, M. A. J. (2018). Relating plant height to demographic rates and extinction vulnerability. Biological Conservation, 220, 104-111. doi:

Hilbers, J., Schipper, A., Hendriks, A., Verones, F., Pereira, H., & Huijbregts, M. (2016). An allometric approach to quantify the extinction vulnerability of birds and mammals. Ecology, 97(3), 615-626.

IUCN. (2017). IUCN red list of threatened species. Version 2017.1.

Kattge, J., Diaz, S., Lavorel, S., Prentice, I. C., Leadley, P., Bönisch, G., . . . Wirth, C. (2011). TRY – a global database of plant traits. Glob. Change Biol., 17(9), 2905-2935. doi: 10.1111/j.1365-2486.2011.02451.x

Salguero-Gómez, R., et al. (2015). The COMPADRE Plant Matrix Database: an online repository for plant population dynamics. Journal of Ecology 103, 202-218

Pimm, S. L., & Raven, P. H. (2017). The fate of the world’s plants. Trends in Ecology & Evolution, 32(5), 317-320.

West, G. B., Brown, J. H., & Enquist, B. J. (1999). The fourth dimension of life: fractal geometry and allometric scaling of organisms. Science, 284(5420), 1677-1679.

Accounting for clonality in plant demography – a story of an unintended article

The story of this article started already some three years ago at the BES symposium Demography beyond populations organised by Rob Salguero-Gómez and colleagues. Two of us were sitting in the auditorium and listening to talks, when almost simultaneously we had the same idea. “Do clonal species differ in their demographic characteristics? They definitely should, given their completely different way of growth! It must be quite easy to find out; we just merge CLO-PLA (the database of plant clonal traits) and COMPADRE.” Therefore, we started to work enthusiastically on the task even in the auditorium. Upon return from the in many aspects great Symposium, we met with Tomáš and told him of our idea and he said that he had had a very much the same idea for some time as well. The three of us set out to work and in a couple of months, we assembled the core of analyses of demographic characteristics of Central European clonal and non-clonal plant species.

So far, the story of the article had been very much as if narrated by a university PR department, yet here, the problems started to emerge. Results of our analyses indicated several effects of clonality and shoot ontogeny type on demographic characteristics, but taken together they did not enable us to come up with a coherent discussion of our paper. Jitka and Zdeněk started to suspect that there must have been at work biases in demographers’ choices of study species. They went through the original works, which the matrix population models (hereafter MPM) in COMPADRE are based on, and found that there are at least three fundamental ways how demographers study demography of clonal species (Fig. 1). Some researchers did not incorporate clonal propagation into MPMs at all. Others treated clonal propagation as a part of mother plant’s growth, while yet others saw clonal propagation as another reproductive pathway similar to generative reproduction. Interestingly and unfortunately as well, there indeed had been a bias in demographers’ decisions (Table 1). Clonal propagation of species with monocyclic shoot ontogeny type was more likely to be neglected, while in species with polycyclic shoot ontogeny type, i.e. overlapping shoot cohorts, clonality was often considered a part of mother plant’s growth. Only clonal propagation of species with monopodial shoot ontogeny type, which typically possess distinct leaf rosettes, was seen as means of reproduction.


At this point, the three of us ended many times in a debate, which of these three approaches is the right one to treat clonal propagation. Finally, during one such heated debate, Tomáš reached out into the bookshelf for now-classical Population biology of plants by J.L. Harper. Harper tackled the problem of defining an individual in the clonal plant’s population by defining a ramet, i.e. such rooted part of the genetic individual, which can live through all life-cycle stages. Reviewing existing demographic studies of clonal species through this prism indicated that in most species clonal propagation needs to be viewed as another mode of reproduction albeit with different genetic and evolutionary consequences. Treating clonal propagation as reproduction does not seem to be a suitable option only in species with very strong integration of ramets (e.g. Carex humilis).

Screen Shot 2017-11-23 at 09.25.06

As we were reviewing the original literature on demography of clonal species, a huge wealth of mainly practical MPM applications unfolded to us and we definitely see this as one of the main assets of MPMs. By drawing recommendations of the ways of clonal propagation incorporation into MPMs, we hope to help to raise the utility of MPMs also for asking general questions about plant demography and evolutionary biology.

Zdeněk Janovský, Jitka Klimešová and Tomáš Herben


The resulting publication of this initiative can be found here:

Janovský, Klimešová & Herben. Accounting for clonality in comparative plant demography growth or reproduction? Folia Geobotanica 1-10



Screen Shot 2017-11-23 at 09.26.19

We got the grant!

On August 5, the COMPADRE/COMADRE team was awarded an NSF grant to further develop our matrix databases. The funded project, “An Open-Access Global Repository of Plant and Animal Demographic Data”, will be led by Judy Che-Castaldo at Lincoln Park Zoo in Chicago, IL. This funding comes from the Advances in Biological Informatics program and will increase the function of the database and make it more user-friendly.


Celebrating our grant success at the ESA meeting in Portland, OR

There are three main parts to the funded project. In the first part, we will finish transferring our data into a relational database, which will run more efficiently and be less error-prone than our old system of spreadsheets. A second part will be to build a data-entry portal that our digitization team will use, so that the data input will be consistent across our digitization nodes around the world. Down the line, this portal will be opened to other researchers who can then contribute their own matrix data. The third part of the project will be to refresh our database website to make it more accessible to a wide range of users, including researchers, teachers, students, and conservation managers.

In addition to improvements to the database itself, we will also be bringing on board a project coordinator who will oversee data digitization and communication across all of our participating nodes. Together, we will develop educational materials and hold user engagement workshops at several scientific conferences each year to spread the word and encourage even greater use of our demographic matrix data for research and in classrooms.

We are so excited about this next step in the COMPADRE/COMADRE project! We hope you will follow along and give us your feedback as we continue to make our databases better and more useful for you.

By Judy Che-Castaldo

Our upcoming workshop in Portland, OR

Over the last few years we have run numerous workshops on using the COMPADRE Plant Matrix Database and COMADRE Animal Matrix Database, and on matrix population models (MPMs) more generally.

Where better to run our next workshop than the upcoming Ecological Society of America (ESA) meeting in Portland, Oregon?

This yearly conference brings together academics, students, and practitioners for a few days of talks and workshops on ecology and allied fields. Attendance is in the thousands — the last time it was in Portland (2012) the meeting drew an amazing crowd of 5000! Although not all attendees will have an interest in MPMs (shame!), there are sure to be more than a handful who’d like to know more.

To help with this, this year we are running a half-day workshop entitled “Introduction to Matrix Population Models and Comparative Population Biology Using the COM(P)ADRE Matrix Databases“.

Drawing from experience garnered over the last few years we will take attendees on a five hour journey from the very basics of matrix modelling to comparative MPM analysis using R. The expert instructors are drawn from the COM(P)ADRE committees and include Owen Jones (Uni Southern Denmark), Roberto Salguero-Gomez (Uni Oxford), Judy Che-Castaldo (Lincoln Park Zoo) and Iain Stott (Uni Southern Denmark).

ESA 2017 attendees were given an opportunity to book for the workshop when they registered for the main conference. However, it should be possible to register as a last minute attendee on-site.

If you can’t make it this time, rest assured that we will continue to run similar workshops regularly at relevant meetings/conferences. We also run them on request: We blogged about one of those here.

Here’s looking forward to some matrix modeling fun in a few days!