Are these interesting results? If yes, worthy of publication?

Memento mori

and remember that all monitoring programs end at some point


I don't remember exactly when I had the idea I describe in the following, but I clearly remember I had a conversation about it with Ryan Chisholm (now at the National University of Singapore) while attending the SIAM meeting in San Diego in the summer of 2013. Marc Mangel was invited, but he could not go. The organizer of the special session on spatial ecology asked Marc if he had a postdoc to recommend as a substitute for him and he recommended me. Semi-surreal experience with only the five people scheduled to give talks present in the room and a couple of them were busy checking emails. I had a very good time, though.

As a side note, I recently had a look at the slides of the talk I gave at SIAM 2013, and there are at least a couple of ideas that could become papers, one on density dependent growth at different spatial scales (which I had partially developed Vincenzi, S. et al. 2010 Detection of density-dependent growth at two spatial scales in marble trout (Salmo marmoratus) populations. Ecol. Freshw. Fish 19, 338–347) and one on rapid genetic differentiation in fish populations living in fragmented habitats. Ryan also talked about his recent (at the time) trip to Cuba and he convinced me - it was not too hard - to visit the island. I went to Cuba for the first time in November 2013, I fell in love with La Habana, and I have been there other 2 times, the last one (October-November 2016) also presenting my research at the Centro de Investigaciones Marinas of Universidad de La Habana.

In La Habana at the beginning of November 2016, humidity at 150% and not too happy about it.

The main idea is this: most monitoring programs start with no end in sight (funding is largely determining the duration of the program) and often with different goals that require vastly different amount of data, for instance estimating average survival or growth rates may require 3 to 5 years of data for salmonids (my model system), while estimating the effects on population dynamics or life-history traits of extreme events such as floods, fires, drought, hurricanes, tautologically require at least the occurrence of one extreme event across multiple populations or multiple rare events in one population (otherwise it is not possible to test hypotheses sensu Platt, J. R. 1964 Strong inferences. Science 146, 347–353.)

How can we define an event as extreme? As I wrote in a 2014 paper (Vincenzi, S. 2014 Extinction risk and eco-evolutionary dynamics in a variable environment with increasing frequency of extreme events. J. R. Soc. Interface 11, 20140441), "[...] extreme events may be defined in terms of extreme values of a continuous variable on the basis of the available climate record (e.g. temperature, precipitation levels) or in the form of a discrete (point) perturbation, such as a hurricane or a heavy storm. This latter category also includes environmental extremes such as unusually big fires, aseasonal floods or rain-on-snow events.". Extremeness includes and goes beyond rarity, and we can assume extreme events to be defined as extreme have recurrence intervals that are longer than the generation time of the species we are investigating.

It follows that if we are interested in the effects of rare events, the minimum duration of the monitoring program is dictated by the expected recurrence interval of the extreme event, but if in absence of extreme events (that is, extreme events are not expected for that species/population) we want to estimate "reliable" average survival and growth rates (which along with recruitment represent the main axes of variation in species' vitals rates and are essential for the development of any self-respecting model of population dynamics) is unclear for how long we need to go on with the monitoring program. I am not the first one to think about this problem in conservation biology, see for instance Chapter 23 of "Design and Analysis of Long-term Ecological Monitoring Studies" (many editors): "Choosing among long-term ecological monitoring programs and knowing when to stop" by Hugh P. Possingham, Richard A. Fuller, and Liana N. Joseph, or Gerber, L. R., M. Beger, M. A. McCarthy, et al. 2005. A theory for optimal monitoring of marine reserves. Ecol Lett 8, 829–837.

In practical terms, I was interested in estimating how stable were the estimates of average, maximum, and minimum (for a sampling interval) survival rates, and somatic growth rates (i.e., body growth) over time in populations of marble, brown, and rainbow trout living in Western Slovenian streams. Briefly, two populations of marble, one of brown, and one of rainbow trout were sampled bi-annually (June and September) between 2004 and 2014 (monitoring is still ongoing, but for various reasons in this study I only included data up to 2014). Those populations were not affected by extreme events since the start of the monitoring program. It is tricky to assess what's "stable" since we estimate rates or probabilities, thus we inevitably have uncertainty in the estimation, but let's put aside this thorny problem for a moment with the assumption we can at least qualitatively see or define "by look" what's stable or not.

Without getting deep into the gory details (and there are many), what I did was to fit capture-recapture models (fish were tagged when bigger than 115 mm) of survival and growth (using von Bertalanffy's growth models) with data available after 3, 4, 5 (and so on) years since the start of the monitoring program. For growth, I used two von Bertalanffy's models, one with random effects (see Vincenzi, S. et al. 2014 Determining individual variation in growth and its implication for life-history and population processes using the Empirical Bayes method. PLoS Comput. Biol. 10, e1003828) and one "classic", that is without individual random effects (I simply used the nls function in R, thus each data point is - wrongly - assumed to be coming from different individuals). For survival, I fitted models Phi(~1) (constant survival) and Phi(~time) (survival varying for each sampling occasion) with different models of probability of capture. With Phi(~time) models, I had the goal to estimate the maximum and minimum survival probabilities over sampling intervals (apart from statistical noise or use of different models of probability of capture, over time maximum survival probability for a sampling interval cannot become smaller and minimum survival probability cannot become bigger).

These are the results for growth, I took asymptotic length (of the average fish) as estimated response variable (estimates with classic nonlinear regression (dot symbol) and estimates with the model with random effects (triangle)), x-axis is the last year of sampling (for, say, 2006, I only keep the data for up to 2006, for 2010 up to 2010, and every time I re-fit models with the corresponding datasets). With the model with random effects, it takes 2 to 4 years to obtain a stable estimate. With the classic method, we need more years, and in some cases, the model gives wrong estimates, since outliers strongly pull the average growth trajectories in their direction (as I also discuss in the PLoSCompBio paper cited above and in another one currently under review). Vertical lines are standard errors of the estimate.

Estimates of asymptotic length using the classic method (dots) and the random-effects model (triangles). The size of symbols is proportional to sample size. LIdri_MT = marble trout in Lower Idrijca, UIdri_MT = marble trout in Upper Idrijca, LIdri_RT = rainbow trout in Lower Idrijca, UVol_BT = brown trout in Upper Volaja.

As I wrote in another paper currently under review: "The vBGF parameters can seldom be interpreted separately, especially when only a few older fish are measured; it follows that the analysis of the whole growth trajectories is necessary for understanding growth variation among individuals and cohorts". The second plot for growth shows the average growth trajectories using the random-effect models using data up to 2006 or up to 2014. Within populations and for all intents and purposes, the average growth trajectories are basically the same.

Average growth trajectories for the 4 salmonid populations estimated with data up to 2006 or up to 2014 with the random-effects model. LIdri_MT = marble trout in Lower Idrijca, UIdri_MT = marble trout in Upper Idrijca, LIdri_RT = rainbow trout in Lower Idrijca, UVol_BT = brown trout in Upper Volaja.

The last plot is about survival. x-axis is again the last year of simulated sampling, dot is the estimate of average survival, up triangle is the maximum estimated survival for a sampling interval, down triangle is the minimum estimated survival for a sampling interval. The estimate of average survival is very stable after just 2 or 3 years (4 to 6 sampling occasions), except for rainbow trout since the small sample size makes the estimate vary quite a bit, as expected.

Dots are average survival probabilities (Phi(~1)), up and down triangles are maximum and minimum survival for sampling occasion (max and min estimates of Phi(~time)). LIdri_MT = marble trout in Lower Idrijca, UIdri_MT = marble trout in Upper Idrijca, LIdri_RT = rainbow trout in Lower Idrijca, UVol_BT = brown trout in Upper Volaja.

The main point and message to take home from this work are that in salmonid populations 2 or 3 years of data may be enough to get "stable" estimates of survival and average growth trajectories, the latter only if using a random-effects model of growth, but > 5 years are needed to estimate maximum and minimum survival probabilities in absence of extreme events (and potentially longer for minimum survival probabilities, as it can be seen in the panels for LIdri_MT and UIdri_MT).

Interesting, surprising, useful, worthy of publication in a conservation journal such as Biological Conservation, Animal Conservation, Conservation Biology?


Posts on my academic life since 2004

Since my formal academic career is likely to end in a few months, I had this idea while on a car trip (San Francisco Airport to Santa Cruz, if you are interested in geographic details) of writing about my academic and research experience since 2004, the year I started my PhD in Theoretical Ecology at the University of Parma, Italy. I believe my experience has been original and interesting, and despite not having been successful in securing a permanent position (or maybe because of it), I am confident I can share valuable insights for young researchers or people outside of academia who want to understand how things work, or might work better. I also have some beautiful pictures to share.

Some brief biographical details. I am 37 years old, I do research in mathematical biology, ecology, and evolutionary biology, I won international grants and awards (in particular a Marie Curie International Outgoing Fellowship, which I started in January 2013 and finished in December 2015), I recently toured South America giving talks on my research in mathematical biology, and I have published so far 30 peer-reviewed papers (43 in total, 3 others in which I am the first author are currently under review, other 2 I may or may not write), and almost one year ago (April 25th 2016, a date that is easy to remember since it is the day in Italy in which the defeat of nazi-fascism is celebrated and remembered, la festa della Liberazione) I received a self-sponsored US Green Card for extraordinary abilities in the Sciences. I live in Santa Cruz, CA, and I have been residing here since June 2010, with a six-month interruption between December 2010 and June 2011. I have loved very much studying, learning how to think and communicate, carrying out my research, giving talks, teaching, traveling the world.

New paper accepted for publication by PRSB

A new manuscript I wrote with my long-term colleagues Marc Mangel, Dusan Jesensek, Carlos Garza, and Alain Crivelli has just been accepted by the Proceedings of the Royal Society B.

This is the cover letter, in which I explain the high-level picture and some details of the analyses (it may help some junior scientists still struggling with cover letters, too much, too little, too many details or not enough etc.).

Cover letter

Santa Cruz, 09/26/2016

Dear Editor,

We are pleased to submit our manuscript “Genetic and life-history consequences of extreme climate events” to PRSB.

The climate change-induced increased frequency and intensity of extreme climate events is one the major threats to the persistence of species. However, when dealing with extreme events, finding the right model system, posing and testing tractable hypotheses on their demographic, genetic, and life-history consequences, and developing an overarching predictive framework is very challenging. First, climate extremes are rare events, and as a consequence most of the empirical studies on their effects have been opportunistic and anecdotal. Then, the demographic, genetic, and life-history effects of extreme climate events are not easily predictable or generalizable across species or habitats, especially when the investigations are not guided by ecological and evolutionary biology theory.

In this work, we test – to our knowledge for the first time - theoretical predictions on the demographic, genetic, and life-history effects of extreme climate events on two populations of a fish species. The two populations have been drastically reduced in size by flash floods that occurred in 2007 and 2009. We used a statistically sophisticated approach that included reconstruction of pedigrees using long-term tag-recapture data (1995 to 2014 from one population, 2006 to 2014 for the other) and genotypes of more than 1,800 unique fish. In particular, we tested for faster life histories, higher variance in reproductive success, and loss of genetic variation after the extreme climate events.

We are confident that our study significantly advances our understanding of the demographic, genetic, and life-history effects of extreme climate events on natural populations and would be of great interest to a broad audience of biologists.

And here below is the abstract (I will soon post the pdf in the Publications page of the website).


Climate change is predicted to increase the frequency and intensity of extreme climate events. Tests on empirical data of theory-based predictions on the consequences of extreme climate events are thus necessary to understand the adaptive potential of species and the overarching risks associated with all aspects of climate change. We tested predictions on the genetic and life-history consequences of extreme climate events in two populations of marble trout Salmo marmoratus that have experienced severe demographic bottlenecks due to flash floods. We combined long-term field and genotyping data, and pedigree reconstruction in a theory-based framework. Our results show that after flash floods, reproduction occurred at a younger age in one population. In both populations, we found the highest reproductive variance in the first cohort born after the floods due to a combination of fewer parents and higher early survival of offspring. A small number of parents allowed for demographic recovery after the floods, but the genetic bottleneck further reduced genetic diversity in both populations. Our results also elucidate some of the mechanism responsible for a greater prevalence of faster life histories after the extreme event.

Further considerations (some self-congratulatory)

These are some thoughts that I shared with one of my colleagues via email before submitting a revised version of the manuscript.

"Brief thoughts. This paper is an example of interdisciplinary work. There is solid life-history theory, we built up from previous work thus giving the sense of solid foundations and a on-going narrative, some hard tests as envisioned by Platt (age at reproduction decreases or not after the floods), demography, classic genetics, and state-of-the-art pedigree reconstruction.

Let's hope it gets accepted as is and we can then congratulate ourselves on an excellent, original work I am very proud of."

New pre-print

Pre-print of my last work is on biorxiv

Vital rates, source-sink dynamics, and type of competition in congeneric species

Simone Vincenzi, Dusan Jesensek, Alain J Crivelli


The estimation of vital rates and life-history traits and how they vary with habitat and population factors are central for our understanding of population dynamics, risk of extinction, and evolution of traits in natural populations. We used long-term tag-recapture data and novel statistical and modeling techniques to investigate how population and environmental factors determine variation in vital rates and population dynamics in the population of brown trout Salmo trutta L. of Upper Volaja (Western Slovenia). Alien brown trout were introduced in the stream in the 1920s and the population has been self-sustaining since then. The population of Upper Volaja has been the subject of a monitoring program that started in 2004 and is currently on going. Upper Volaja is also a sink, receiving individuals from a source population living above an impassable waterfall. We estimated the contribution of the source population on the sink population and tested the effects of temperature, population density, and early environment on variation in vital rates and life-history traits among more than 4,000 individually tagged brown trout that have been sampled since 2004. We found that fish migrating from the source population (>30% of population size) help maintain high population densities despite poor recruitment. Neither variation in density nor in temperature explained variation in survival or growth; the best model of survival for individuals older than juveniles included cohort and time effects. Fast growth of older cohorts and higher population densities in 2004-2006 suggest very low densities in early 2000s, probably due to a flood event that caused a strong reduction in population size. Higher population densities, smaller variation in growth and weaker maintenance of size hierarchies with respect to endemic marble trout suggest that exploitative competition for food is at work in brown trout and interference competition for space is operating in marble trout.

Data is here

I will provide the R code soon

Floods causing demographic and genetic bottlenecks

I am currently working on testing hypotheses on the effects of flood events causing massive mortalities on reproduction, survival, and growth of marble trout; more in general, I am testing the effects of massive mortality events on the adoption/evolution of life histories.

Here is a schematic representation of the model system and the processes I am investigating that I have been using in some of my recent talks, more a narrative than a paper Introduction.

Tagging and sampling of marble trout populations started in 1996 and 1998 for Zakojska and Gacnik (with the introduction of the parental cohorts in the streams, see previous post), respectively, in 2002 in Huda, in 2004 in Lower Idrjica and Upper Idrijca, and in 2006 in Studenc, Lipovesck, Trebuscica, and in 2008 in Svenica. Apart from Zakojska and Gacnik, in all the other populations the estimation of density started years before the start of the tagging study.


We estimated density of fish older than 0+ (i.e. fish that survived the first winter) using a two-pass removal at each sampling occasion.

Let's have a look at the population dynamics of two marble trout populations. The population of Lipovesck was doing fine after the start of the monitoring program...


but between 2007 and 2008 there was an episode of massive mortality.


Also the population of Zakojska was doing fine...


but also in this case we observed an episode of massive mortality between 2007 and 2008.


It turns out that the streams in which marble trout live are quite mellow and peaceful most of the time, but sometime they are affected by flash floods and debris flows. Flash floods and debris flow occur in the area in which marble trout live causing substantial damages to infrastructures, and death or downstream displacement of trout. Flash floods are characterized by very short time scales (less than a few hours), stream discharge quickly goes up and quickly goes down, and mean daily water discharge may be a very poor indicator of the occurrence of a flash flood. The area in which marble trout live receives more than two times the average annual Slovenian rainfall and it is one of the wettest regions of Europe. Flash flood in Western Slovenia tend to occur from September to November, but in the last few years they also occurred in the Spring.

Flood pic

At this point you might want to know what happened to the populations of Lipovesck and Zakojska.

The population of Lipovesck was affected by another flood event in 2009, but it recovered thanks to a massive production of young in 2011. The population of Zakojska is on the way back, but the surviving fish were located only in two section of the stream, while before the flood the entire stream was inhabited by marble trout; thus, the recovery is expected to be slower than in Lipovesck.


Topography also plays a role. In Lipovesck, we were monitoring and sampling only the lower part of the stream,  but above a 1.5 to 2 m-tall waterfall there were fish alive after the flood that contributed to population recovery (e.g., most of the young born in 2011 were probably produced upstream).


In Zakojska, some sectors are isolated (fish can move down, but not up). Fish can move up only from sector 4 to 5, and from 1 to 2. After the flood (in 2008, 2009) fish were alive and present only in sector 5 (thus sector 6 is lost since there are no chances of spontaneous colonization - no fish above sector 6) and in sector D (downstream). Numbers above sectors in the figure below are point estimates of annual probability of moving from one sector to another (data pre-2008, estimated using MARK).


What happens when a flash flood or debris flow occur?

This is an illustration of a population bottleneck. Each marble is an individual or a group of individuals in a population.


At some point, an environmental extreme event such as an earthquake, flood, fire, drought, cause a sharp reduction in population size, i.e. a population bottleneck. Just a few individuals are able to pass through the bottleneck. This event has demographic consequences by reducing the number of individual alive and genetic consequences by very likely reducing the genetic diversity of the population, since just a fraction of the original genetic diversity is expected to be present after the population collapse.


If there are no individuals left, or the few that survived are not able to produce a strong-enough cohort of young, that population is likely to go extinct.


In other cases, the individuals passing through are the among the most fit individuals in the population or have particular traits that help them reproduce successfully or are just able to reproduce and produce a strong-enough cohort of young by chance.

Bott4In that case, the population bounces back to safe abundances, as in the case of Lipovesck.



I am currently finalizing the pedigree reconstruction for the populations of Lipovesck and Zakojska, using 77 and 94 population specific SNPs, respectively. The use of SNPs was necessary as marble trout populations present low to absurdly low genetic diversity (for one population - Huda - we almost found no "usable" polymorphisms" despite multiple DD-RAD runs). In particular,  I am testing differences in allelic diversity before and after the flood (expected a decrease after the flood), differences in life histories (expected faster growth and younger age at first reproduction after the flood), differences in survival (expected higher survival after the flood due to relaxation of density-dependent pressure). The hypothesis-testing analyses will be accompanied by analyses that are more exploratory in nature, such as traits of successful reproductors, movement of young after the flood, variance in parents per offspring,

I will soon post updates on the study.

Correlation between stream temperatures in Slovenian streams in which marble trout live

This is an update on my research and I will try to post more often in these last months of my Marie Curie Fellowship. Files are hosted on my github page. Data have been collected by Alain Crivelli and Dušan Jesenšek since 1996. Some info on marble trout, the conservation program, and Western Slovenian streams here below.

1. Marble trout and Western Slovenian streams

Marble trout is a freshwater resident salmonid endemic in the Adriatic basin of Slovenia. Whether there are still pure marble trout populations living in the Po river system (Northern Italy) is subject of current research. Marble trout live in streams with mean summer temperature below 15°C and winter temperature ranging from 0 to 5 °C. Marble trout spawn in November-December and offspring emerge in May-June.
The Marble Trout Conservation Program started in 1993 in the upper reaches of the Soca River basin and its tributaries - the Idrijca and Baca Rivers - in Western Slovenia. Eight pure marble trout populations, all isolated and separated from the downstream hybrid marble-brown trout zone by impassable waterfalls, live in headwater streams in the basins of Soca, Baca, and Idrijca Rivers: Huda, Lower Idrijca, Upper Idrijca (in the map below Lower and Upper Idrijca are grouped together), Lipovesck, Studenc, Svenica, Zadlascica, Trebuscica.
Other two populations (Zakojska and Gacnik) have been created by translocating the progeny of the Zadlascica (Zakojska) and Trebuscica X Lipovesck (Gacnik) in 1996 and 1998, respectively.


2. Analyses

For some of the analyses I intended to carry out (temperature-dependent survival, growth, and recruitment), it was necessary to have complete temperature records for all streams since the start of the sampling. However, there were some missing data (sometimes whole seasons/years) in evert stream. The temperature .csv files are (stream_name)_temp.csv, the first column is the Date, the second is the mean daily temperature (Temp). Start by sourcing the file Temp.r, which is reading all the temperature files and merging them together (r scripts are here).


The output temp.all.df (along with the production of a ten-panel plot with stream-specific monthly temperature boxplots for 2009-2013) is a data.frame with columns Date, Temp, Year, Month, Stream, Calc (Meas = temperature has been recorded, see below for other values) (see below).


Daily Water Temperature (C)
Date Temp Year Month Stream Calc
1996-07-04 10.98 1996 7 Zak Meas
1996-07-05 10.99 1996 7 Zak Meas
1996-07-06 11.26 1996 7 Zak Meas
1996-07-07 11.19 1996 7 Zak Meas
1996-07-08 11.06 1996 7 Zak Meas
1996-07-09 9.96 1996 7 Zak Meas
1996-07-10 9.85 1996 7 Zak Meas
1996-07-11 10.07 1996 7 Zak Meas
1996-07-12 10.51 1996 7 Zak Meas
1996-07-13 11.08 1996 7 Zak Meas

Then, I tested the correlation between stream temperatures between pair of streams (one is the target - the one with missing data - and the other is the tested). I used the temperature data of the tested stream with the highest correlation with the temperature data of the target stream to impute the missing temperature data in the tested stream.
The Temp.corr.f function tests the correlation between water temperature data recorded in different streams.

source("Temp.corr.r") # contains Temp.corr.f
Temp.tb = Temp.corr.f(temp.all.df)

The Temp.tb data.frame has columns target stream (tar), tested stream (var), correlation between stream temperature of the two streams (cor), years with common number of days with temperature recorded (common.years), years with missing data for the target stream (miss.years), and years with missing data for the target stream, but with complete data for the tested stream ( The years in can thus be used to impute the missing data.
The correlation between water temperature of streams are typically very high (mean correlation[sd] = 0.97[0.01]).

Correlation of water temperature
tar var cor years.cor common.years miss.years
Zak Gac 0.95 5 2001-2002-2005-2009-2013 1996-1997-1999-2000-2006-2008-2010-2011-2014 2006-2008-2010-2011
Zak Sve 0.95 3 2002-2009-2013 1996-1997-1999-2000-2006-2008-2010-2011-2014 2006-2008-2010-2011
Zak Stu 0.97 3 2005-2009-2013 1996-1997-1999-2000-2006-2008-2010-2011-2014 2006-2008-2010-2011
Zak LIdri 0.97 3 2005-2009-2013 1996-1997-1999-2000-2006-2008-2010-2011-2014 2006-2008-2010-2011
Zak UIdri 0.96 3 2003-2005-2013 1996-1997-1999-2000-2006-2008-2010-2011-2014 2006-2010-2011

In each stream, I imputed the missing data (1) using the temperature data from the tested stream with the highest correlation with the target stream and (2) by applying the best model (linear or non-linear - gam -, chosen according to best prediction) linking the water temperature data of the two streams. The r script for imputing missing data is in Temp.filling.r.


The output of the script is the data frame temp.all.df with columns: Date, Temp, Year, Month, Stream,
Calc (Meas = temperature recorded in the stream, Gac2005 = in one year (1997) we had missing data for Gac and the only acceptable data for imputing was coming from Gac in 2005, Same_as_a = Same temperature as days after (just for a few days missing), Same_as_b = Same temperature as days before (just for a few days missing), Zak2012 = in one year (1997) we had missing data for Zak and the only acceptable data for imputing was coming from Zak in 2012, Stream_name = stream whose temperature data was used to impute missing data, degree_days = degree days for the day using 5C as base temperature, Sampling_Season = Summer for June, July, September - Winter for the rest of the year). Sampling occurred either in June or September or in both.

Daily Water Temperature (C)
Date Temp Year Month Stream Calc degree_days Sampling_Season
2006-08-19 13 2006 8 Stu Meas 7.8 Summer
2006-08-20 13 2006 8 Stu Meas 7.8 Summer
2006-08-21 13 2006 8 Stu Meas 8.1 Summer
2006-08-22 13 2006 8 Stu Meas 7.6 Summer
2006-08-23 13 2006 8 Stu Meas 7.6 Summer
2006-08-24 12 2006 8 Stu Meas 7.3 Summer

Temperature data is now ready to be used to test differences in water temperature between streams, and temperature-dependent survival, growth, and recruitment.

Slides of a recent talk I gave at the Hopkins Marine Station (De Leo's lab)

Slides of a recent talk I gave at the Hopkins Marine Station; I was invited by my former PhD supervisor Giulio De Leo.


Eco-evolutionary responses to extreme events

Reference paper

Extinction risk and eco-evolutionary dynamics in a variable environment with increasing frequency of extreme events

Slides (some colors are missing for unexplained reasons)

Paper published and media coverage

My colleagues and I (Simone Vincenzi, Scott Hatch, Thomas Merkling, Alexander S. Kitaysky) recently published the paper "Carry-over effects of food supplementation on recruitment and breeding performance of long-lived seabirds" in the Proceedings of the Royal Society Biological Sciences (you can find the un-gated paper here).

It has been a long and challenging work, from data preparation to multiple manuscript revision, but it has been worthwhile, as results are of general interest and intriguing.

From the Cover Letter we sent to PRSB's Editor: "This is the first experimental test of the long-term effects of controlled variation in early food availability in long-lived wild animals. In addition to casting light on some of the ecological consequences of variation in early food availability, our results also have pivotal consequences for conservation science".

------ Some other excerpts from the cover letter here below

The supplementation of food for wild animals is extensively applied as a conservation tool to increase the local production of young, but the effects of such food supplementation on the subsequent recruitment of long-lived animals into natal populations are largely unknown. For long-lived species, studies are generally observational due to the long time periods required for individuals to reproduce and/or complete their life cycles (SV note: I am lucky enough to work on two model system with tagging that started in both cases in 1996). Our experimental study, more than a decade long, of the long-term effects of early food supplementation on long-term performance of a long-lived species is thus a novel, original, and exciting contribution. We used the unique experimental system of kittiwakes breeding on Middleton Island (Alaska) to test the alternative hypotheses that food supplementation early in life (a) increases overall fitness of birds, or (b) delays viability selection, with no consequences for the long-term dynamics of the species.

The results of our study are exciting and surprising. Through rigorous statistical and modeling analyses, we found that delayed viability selection is decreasing the recruitment rate of food-supplemented chicks with respect to control birds. We also identified a potential mechanism for the delayed viability selection, i.e. more intensive brood reduction in control nests.Lifetime reproductive success of a subset of kittiwakes that thus far had completed their life cycle was not affected by the food supplementation during development. However, per-nest contribution of recruits was still higher for food-supplemented nests due to their greater productivity compared to control nests, thus suggesting a positive net effect of food supplementation on recruitment.


The paper received media coverage on ScienceNews with a very clear and accurate article written by Sarah Zielinski in the Wild Things blog

Boring, but important, note (as we all know that talk is cheap, but money buys whiskey) about funding:

Fieldwork and modeling were supported by the US Geological Survey and North Pacific Research Board (Project no. 320, BESTBSIERP Projects B74, B67 and B77). S.V. is supported by an IOF Marie Curie Fellowship FP7-PEOPLE-2011-IOF for the project ‘RAPIDEVO’ on rapid evolutionary responses to climate change in natural populations, and by the Center for Stock Assessment Research (CSTAR). The MC Fellowship FP7-PEOPLE-2011-IOF and the Institute of Arctic Biology at UAF provided funds to cover the publication costs.

Manuscript submitted

I recently submitted a new manuscript on vital rates and life histories in marble trout. Dense paper, lots of models, lots of results. Currently under review. Here below are the Title and Abstract.


Within and among-population variation in vital rates and population dynamics in a variable environment --- Vincenzi, Mangel, Jesensek, Garza, Crivelli.


Understanding the causes of within- and among-population differences in vital rates, life histories, and population dynamics is a central topic in ecology. In order to understand how within- and among-population variation emerge, we need long-term studies that include episodic events and contrasting environmental conditions, tag-recapture data for the estimation and characterization of individual and shared variation, and statistical models that can tease apart population-, shared-, and individual contribution to the observed variation.

We used long-term tag-recapture data and novel statistical and modeling techniques to investigate and estimate within- and among-population differences in vital rates, life histories and population dynamics of marble trout Salmo marmoratus, a narrow endemic freshwater salmonid. Only ten populations of pure marble trout still persist in Western Slovenian headwaters. Marble trout populations are also threatened by floods and landslides, which have already caused the extinction of two populations in recent years.

In particular, we estimated and determined causes of variation and trade-offs within- and among populations in growth, survival, and recruitment in response to variation in water temperature, density, sex, early conditions, and extreme events.

In all ten populations, we found that the effects of population density on traits were mostly limited to the early stages of life and that individual growth trajectories were established early in life. We found no clear effects of water temperature on survival and recruitment. Population density was variable over time in all populations, with flash floods and debris flows causing massive mortalities and threatening population persistence. Apart from flood events, variation in population density within streams was largely determined by variation in recruitment, with survival of older fish being relatively constant over time within populations, but substantially different among populations. A fast- to slow-continuum of life histories in marble trout populations seemed to emerge, with slow growth associated with higher survival at the population level, possibly determined by food conditions and age at maturity.

Our work provides unprecedented insight into the causes of variation in vital rates, life histories, and population dynamics in an endemic species that is teetering on the edge of extinction.

Some reflections on my science

Since I had to change the links to my publications due to some obscure passage of pdfs from one folder to another, I had a chance to have a look at all the papers I published so far. 40 total, 28 as first author, 4 under review (2 as first author, other 2 under review). Surprisingly (or not, upon further reflection) I barely remember the content of most of my papers and I have little idea on how they were originally thought, what was the development, what was the contribution of co-authors, why I used certain methods and not others. I saw big tables I did not remember I had prepared. I saw a Figure in which fish are one year older than what they should be (I also thought I sent the correct Figure during the revision process, apparently not). I read long Introductions and longer Discussions (I write a lot, no doubt). I remember long struggles to get papers accepted even if I currently do not remember the major contentious points.

Just to be clear, I do not have any memory disorder. However, I have published in many different areas, in part because I prefer to zig-zag than follow a straight-ish line, in part because I have been supported by soft money throughout all my career and I haven't been too rigid in my research/grant choices. I also tried to use novel methods (for me or in general), since I like to challenge myself and expand my research tools. I tend to go very deep and very fast in my research and this - like cramming for a test - is not conducive to long-term retention of information.

This "discovery" made me think about my research trajectory, what kind of tools and skills I have acquired, and whether production of science is like the production of eggs in fish: you give your contribution and you let it find its way.