DemGenTrout 1.1

No preview image

2 collaborators

Default-person Béatrice Frank (Author)
Philippe Baret (Advisor)

Tags

brown trout 

Tagged by Béatrice Frank almost 6 years ago

fish 

Tagged by Béatrice Frank almost 6 years ago

population dynamics 

Tagged by Béatrice Frank almost 6 years ago

population genetics 

Tagged by Béatrice Frank almost 6 years ago

salmo trutta 

Tagged by Béatrice Frank almost 6 years ago

Child of model DemGenTrout 1.0 preview imageDemGenTrout 1.0 Parent of 2 models: DemGenTrout 1.2 and DemGenTrout 1.3
Visible to everyone | Changeable by the author
Model was written in NetLogo 5.0.2 • Viewed 185 times • Downloaded 16 times • Run 0 times
Download the 'DemGenTrout 1.1' modelDownload this modelEmbed this model

Do you have questions or comments about this model? Ask them here! (You'll first need to log in.)


WHAT IS IT?

This is the version 1.1 of a brown trout demogenetic model developed by Beatrice Frank during her PhD thesis (Supervisor: Prof. Philippe Baret, Earth and Life Institute, Universite catholique de Louvain). In comparison with DemGenTrout 1.0, the initialisation of this version is faster.

HOW IT WORKS

The model description follows the ODD (Overview, Design concepts, Details) protocol (Grimm et al., 2006; 2010). This protocol consists of seven elements. The first three elements provide an overview of the model, by stating (i) the question addressed, (ii) its structure in terms of entities (agents), state variables, temporal and spatial scales, and (iii) its dynamics, i.e., the order in which the processes (submodels) of the model are executed. In the fourth element, general concepts underlying the model's design are explained. The remaining three elements complete the model description by providing further details about initialisation, input data, and submodels.

OVERVIEW

PURPOSE

The model, named DemGenTrout, was designed to understand how anthropogenic disturbances can affect a brown trout (Salmo trutta L.) population living in a river / nursery brook system of the Lesse River network (Belgium). Changes in the demogenetic (i.e., demography and genetics) structure of the population were monitored, and at a latter stage, the model was used to predict how trout populations might respond to migration barriers and stocking with hatchery trout (Frank & Baret, 2013).

ENTITIES, STATE VARIABLES, AND SCALES

Two types of entities are comprised in the model: two streams, representing the studied hydrological system, and trout agents. The stream agent corresponds to the tributary and the main river, and each stream has three state variables: code, current flow, and current water temperature. Trout agents are characterised by 15 state variables: natal stream, sex, genotype, and week of birth (static variables); age, stage, status, length, weight, condition factor, number of offspring, current status of spawning, movement or return, and current stream they are in.

The spatial resolution and extent of the model are defined by the two possible locations for trout in the system: the Chicheron Brook (referred as stream C), and a section of the Lesse River (stream L). One time step represents one week and simulations last 35 years, corresponding to 10 trout generations.

PROCESS OVERVIEW AND SCHEDULING

Nine processes are considered in the model: the update of hydrological conditions for each stream agent; survival, growth, reproduction, ageing, hatching, moving downstream, post-spawning homing, and leaving the system forever for each trout agent.

Each time step, stream agents first update their water temperature and flow. Second, trout agents challenge their survival. If alive, they update their length, weight, and condition factor. Third, if spawning conditions are met, trout possibly produce offspring that will hatch ten weeks later. Fourth, trout update their age and stage one week before the beginning of the hatching period. Fifth, if conditions for movement are met, trout of the tributary have the possibility to move downstream (i.e., from stream C to stream L). Sixth, alive spawners return to their previous stream one week after the spawning process (post-spawning homing behaviour). The last week of the year, some juveniles (i.e., trout of age 1 and 2) living in stream L do not settle in this stream and disappear from the system to search for more suitable locations.

All actions occur in the same predetermined order, and are executed either once per time step, at specific time windows, or only once a year, as specified in the fixed scheduling of the model:

  • Initialise trout and stream agents.
  • Weekly stream action: update stream hydrological conditions (water temperature and flow).
  • Weekly trout actions: determine how many trout die in each stream; update trout length, weight, and condition factor.
  • Trout actions during the spawning time window: identify candidate spawners using physiological criteria; move candidate spawners upstream when hydrological conditions are suitable; produce offspring in each stream.
  • Trout actions during the movement time window: identify candidates for migration using physiological criteria; move candidate migrants downstream when hydrological conditions are suitable.
  • Yearly trout actions: increment trout age and update stage in each stream.
  • Trout action during the hatching time window: reveal offspring in each stream.
  • Yearly trout action: move upswimming spawners back to stream L.
  • Yearly trout action: remove young trout of stream L from the system.

DESIGN CONCEPTS

Basic principles: The model is underpinned by demogenetics, an emerging field in ecology that integrates both population dynamics and population genetics. Population abundance and structure can be influenced by density-dependent or density-independent processes. In the model, we considered both types of processes at the submodel-level: on the one hand, the downstream migration of young trout is density-dependent and is expected to have a strong effect on the regulation of abundance in the brook; on the other hand, environmental variability may affect the behaviour of individuals such as their decision of moving between streams which in turn shapes the demographic structure. Individuals inhabiting large rivers usually move into first-order tributaries to spawn, and return to their original territory once reproduction is complete (i.e., post-spawning homing behaviour). Homing to natal rivers is particularly strong in salmonids. However, some individuals might decide to not reproduce in their home river (i.e., straying behaviour). Natural selection is the key micro-evolutionary process of population genetics considered in the model. Its effect is evaluated indirectly through the use of microsatellite markers, which allow describing the genetic structure of the population with F-statistics.

Emergence: Survival, growth, spawning, and movements of individuals are imposed behaviours that are empirically described in the model using real data. Spawning and downstream movement processes are affected by hydrological variables. Population dynamics and genetics (i.e., demogenetics) emerge from these behaviours acting at the individual level: on the one hand, the demographic structure of the population (fry, juveniles and adults proportions) is shaped by the number of reproducing individuals; on the other hand, the number of produced offspring is determined by the female length and new genotypes emerge as a result of genetic adaptation and natural selection.

Adaptation: Juvenile and mature adult movements are the adaptive traits in the model. They are modelled as indirect fitness-seeking, in which individuals make decisions of life history tactics that indirectly contribute to future success at passing genes on. Juveniles decide whether and when they migrate to the main stream, and this decision affects their growth. Mature adults decide whether, when and where they spawn each year, and this decision affects the offspring production in each stream. The selection of spawners is based on their body condition factor, and offspring length is inherited from their parents.

Objectives: Fitness-seeking is not explicitly considered in the model.

Learning: This concept is not explicitly considered in the model.

Prediction: This concept is not explicitly formalised in the model. However, tacit predictions are included because fitness-seeking is implicitly modelled. In addition, the tributary is supposed to offer more favourable zones for reproduction, so the probability of reproducing successfully is highest if trout chose to spawn in the brook.

Sensing: No sensing mechanisms are explicitly represented. Individuals are assumed to know the status or value of all their state variables, which affect their survival, growth, spawning and movements. They also have access to information about their current stream, such as water temperature and flow.

Interaction: Individuals do not directly interact together, but competition for shared resources is taken into account in the downstream movement process of juveniles. The probability of moving depends on the number of similar age trout present in the tributary (density-dependent population regulation).

Stochasticity: Most processes of the model are drawn from empirical probability distributions in order to include individual variability. The model incorporates environmental variability through the hydrological variables (i.e., input data), for which additional stochasticity was introduced using random floating point numbers. Survival and downstream dispersal are stochastic events. Whether a trout actually dies or moves to the main stream is determined by comparing a random number to the survival rate or the movement probability. Stochasticity is also used in the spawning process for the selection of spawners, the production of offspring and the definition of their state variables.

Collectives: Collectives are not included in the model.

Observation: The following demographic and genetic outputs were recorded to follow the changes through time of the brown trout population demogenetic structure: the number of migrant juveniles and the number of spawners moving between both streams over each year and, for each stream, the total number of individuals, the number of individuals with a body length > 70 mm, the number of offspring, and the genotypes of 48 randomly chosen individuals. Three genetic outputs were derived from genotypes using the adegenet and pegas R packages: (i) the effective number of alleles, EA, to measure the genetic diversity for trout of each stream, (ii) the inbreeding coefficient, Fis, to measure the extent of genetic inbreeding within trout of each stream, and (iii) the fixation index, Fst, to measure the degree of genetic differentiation between trout of both streams.

DETAILS

INITIALISATION

Initialisation does not vary among simulations. We used a seed value of 1223251200 to start the random number generator. Each simulated year starts on October 1, and lasts 52 weeks. Hydrological variables (water temperature and flow) are assumed uniform throughout a stream and are initialised with the input data corresponding to the first week of simulation (see next section Input data). Trout abundance, age-class repartition, age-specific length and condition factor random Normal distributions in both streams, as well as age-specific proportions of trout living in stream L that changed their natal-stream state variable from L to C are initialised with the text file ''init-trout''.

In the model, length distributions are truncated at +- 4 standard deviations. Values of initial trout abundance and age-class repartition of individuals in each stream were calibrated using standard genetic algorithms implemented in BehaviorSearch. Trout condition factor corresponds to Fulton's coefficient K, which assumes that the standard weight W of a fish is proportional to the cube of its length L, i.e., K = W / L^3. The value of this factor is 1 when an individual has a healthy weight for its length; a value > 1 indicates an overweight, and a value < 1 a poor condition. Trout current and natal stream are assigned according to the location of the individual or to its stream of birth. As the presence of natal homing behaviour was demonstrated for trout of the studied system, some randomly selected individuals living in stream L change their natal-stream state variable from L to C. Their proportion was obtained by calibration.

The other trout variables are initialised as follows: each trout weight is calculated as its body condition factor times its length cubed according to Fulton's formula; trout sex is randomly assigned with even probability of being female or male; each trout genotype is constituted of twelve loci, with two alleles at each locus randomly assigned in accordance with observed allelic frequencies (see text files ''input-genotypes-streamC-12markers'', ''input-genotypes-streamL-12markers'', ''input-genofreq-streamC-12markers'' and ''input-genofreq-streamL-12markers''); age of adults is drawn from a uniform integer distribution between 3 and 6; stage is set according to age; trout status is set to 'non-spawner'; the week of birth is set to the trout age multiplied by -52; the number of offspring is set to 0; and the Boolean variables spawned?, moved-to-spawn? and returned? are set to false.

INPUT DATA

The input data of the DemGenTrout model are weekly time series of observations: water temperatures (C) and flow rates (m/s) in both streams (see text file ''input-data''). The first four years are real field data (2004-2008) and the 31 subsequent years were simulated using the Holt-Winters method implemented in R.

SUBMODELS

The model contains 51 parameters, which are listed with their values in the text file ''init-params''. The values of 38 of them were directly obtained from field studies conducted on the Lesse River / Chicheron Brook hydrological system. For the length-heritability parameter, the value was derived from the literature. Four parameters were guestimated (i.e., we had only an idea of what would be a realistic value for each of them) and 8 parameters were calibrated.

The model comprises nine submodels or processes, and their description follows the fixed scheduling presented in the Process Overview and Scheduling section. We used the following conventions when a parameter name is referring to either both streams or several age classes: streamX, where X corresponds to C or L; ageY, where Y corresponds to age class 0, 1, 2 or >= 3; ageZ, where Z corresponds to age class 1 or 2.

1. Update stream hydrological conditions: Each week, the flow and water temperature of each stream are updated from the input data. They are assumed uniform throughout a stream.

2. Kill trout in each stream: Each week, a random number between 0 and 1 is drawn for each trout. If this number is higher than the survival rate corresponding to the trout current age and stream location (streamX-ageY-survival parameters), then the trout dies. Probability density functions were used to get annual trout survival rates, which were transformed into weekly values in the model (i.e., survival rates exponent 1/52). The value of streamC-age3-survival was guestimated from streamL-age3-survival. Three mortality sources were indirectly taken into account when the survival functions were drawn: predation by spawners for young trout of the brook during the reproduction period, high temperatures in summer for all trout, predation by grey herons (Ardea cinerea L.) in autumn for trout of age equal or higher than 2. For brown trout, the upper lethal temperature in fresh water is at 24.7_C. Trout frequently eat eggs or smaller conspecifics, while birds have been observed to remove large numbers of individuals from small shallow streams. As a high mortality was observed in the brook for spawners during winter, mainly due to predation by herons, the survival rate of trout moving from stream L to stream C for reproduction (i.e., individuals for which the moved-to-spawn? state variable is true) was divided by 2 (predation-factor; guestimated value).

3. Update trout length, weight and body condition factor in each stream: Growth in length is modelled with the von Bertalanffy equation, for which the parameters differ between both streams. Each week, each trout length is updated according to its previous length: L(t+1) = L(t) + k * (Linf - L(t)), where k is the growth coefficient (streamX-parK parameter) and Linf is the asymptotic length (streamX-max-length). Then, the trout new length is used to calculate its healthy weight Wh, using parameters a and b of the length-weight relationship (streamX-parA and streamX-parB), which also vary between streams: Wh = a * L(t+1)^b. Finally, the weight W of each trout is computed as its relative condition factor (Kr; corresponding to the condition-factor state variable) multiplied by its healthy weight according to Le Cren's formula: W = Kr * Wh. An inter-individual variability is randomly generated for the condition factor, which value can vary by +- 0.008 (a conditional statement avoids values strictly lower than 0.5).

4. Reproduce trout in each stream: Trout are iteroparous; spawning only occurs during the breeding season each year, but individuals can spawn several times during their life. Trout adapt their reproductive behaviour to external conditions and their own state, and the location and time of spawning are influenced by both their physical state and hydrological conditions. Each week of the spawning period (i.e., week 6 to 17; spawn-start and spawn-end parameters), each trout first determines if it belongs to the group of candidates for reproduction. Then, when hydrological conditions in stream L are suitable, each candidate spawner determines if it moves upstream (i.e., from stream L to stream C). Eventually, candidates of each stream become spawners when they produce offspring.

4.1. Identify candidate spawners: Each trout determines whether it meets a series of criteria: adequate age (equal or higher than 3; spawn-min-age parameter), length (normal distribution of mean and standard deviation equal to spawn-mean-length and spawn-sd-length) and condition factor (normal distribution of mean and standard deviation equal to spawn-mean-cond and spawn-sd-cond), not spawned or moved to spawn the current year. If it does, its status is changed from 'non-spawner' to 'candidate-spawner'. Then, candidate spawners living in stream L are randomly selected to move upstream for reproduction, among which 55% (1 - \moved-prop) of individuals born in stream C (natal-homing behaviour) and 45% (moved-prop) of individuals born in stream L (straying behaviour).

4.2. Move candidate spawners upstream: The upstream movement of candidate spawners occurs when flow rate in stream L follows a log-normal distribution of mean and standard deviation equal to spawn-mean-flow and spawn-sd-flow parameters. On weeks meeting this criterion, the current location of each selected candidate spawner as well as each candidate spawner born in stream C and currently living in stream L is changed from 'L' to 'C'.

4.3. Create offspring: Each week of the spawning period and for each stream, the number of crosses is computed as half the total number of candidate spawners. Females can reproduce only once each year, while several small subordinate males can contribute to the fertilization of the eggs of one female (polygamous mating strategy). For each cross, the female is selected among the list of candidate female spawners ranked in descending order by condition factor. The number of males per female (n) is randomly drawn from a uniform distribution from 1 to 4. Then, the n males are randomly selected among a list containing all the candidate male spawners minus n males having the highest condition factor. Upswimming spawners are prioritized over other spawners of stream C by a statement that first identifies the candidate spawners that have moved (moved-to-spawn? state variable is true), and then selects them before the others.

The number of offspring produced per female per week in each stream depends on its length and follows a linear function, with a slope alpha equal to (F2 - F1) / (L+ - L-) and an intercept beta given by F1 - alpha * L-. Values of L- and L+ (offprod-min-length and offprod-max-length parameters) were derived from observations at the trapping facility; values of alpha and beta were obtained by linear regression of produced offspring vs. female length, and were then used to derive the value of F1 (offprod-min), which was considered identical for both streams. F2 corresponds to offprodX-max and was set to 168 for stream C and 5 for stream L (calibrated values). A demographic explosion was observed when the model was simulated over 35 simulated years. Consequently, we decided to limit offspring production in stream C by dividing offprodC-max by 10 whenever trout abundance in this stream reaches a capacity of 6700 individuals (streamC-capacity parameter).

The process continues until the precomputed number of crosses is reached. After each cross, the num-offspring state variable of females and males that have effectively spawned is updated. The state variables of each offspring are set as follows. Genotype is given by a random combination of the parental genotypes (the female genotype and a random genotype constructed from the genotypes of the n males), with equal probability of getting each parent allele; age and stage are set to -1 and 'fry', respectively; sex is randomly assigned with even probability of being female or male; current and natal streams are assigned according to the location where the reproduction occurs; birth week is set to the current week; body condition factor is drawn from the same stream-specific normal distributions used for model initialisation and weight is calculated as condition factor times length cubed according to Fulton's formula.

Offspring lengths Loff are modelled as being inherited from their parents, following the equation: Loff = mupar + devoff(P) = mupar + devoff(G) + devoff(E), where mupar is the parental population mean, devoff(P) is the total phenotypic deviance of the offspring population, devoff(G) and devoff(E) are the genetic and environmental contributions to the total phenotypic deviance, respectively. First, mupar and varpar are derived from the distribution of the parents lengths at the fry stage, Lpar0, known from the reverse von Bertalanffy growth equation: Lpar(0) = Linf - ((Linf - Lpar(t)) / (1 - k)^t), where k and Linf correspond to the streamX-parK and streamX-max-length parameters, and t is the time elapsed since the birth of the trout (k and Linf are set according to the natal stream). Second, devoff(G) is given by sqrt(h2) * (devpar1 + devpar2), where h2 is the narrow-sense heritability, devpar1 (length-heritability parameter) and devpar2 are the deviation of each parents length from the population mean mupar. Third, devoff(E) is drawn from a random normal distribution with mean 0 and variance (1-h2)2 * varpar. To avoid negative lengths, each Loff value not included in a trout length distribution defined for each stream by the meanl-birth-X and sdl-birth-X initial conditions, truncated at +- 4 standard deviations and rounded to 1 mm when negative values are drawn, are set to a random value drawn from this distribution.

5. Increment age and update stage of trout in each stream: One week before the beginning of the hatching period (i.e., week 15; hatch-start parameter - 1), the age of each trout is incremented by one. The stage is updated as follows: if the age is strictly lower than 1, the stage is set to 'fry'; if the age is equal to 1 or 2, it is set to 'juvenile'; if the age is strictly higher than 2, it is set to 'adult'. All individuals die once they reach the age of 7.

6. Reveal offspring in each stream: Offspring that have been previously created during reproduction have an age equal to -1 and are hidden until a 10-week delay is reached. When it does, offspring progressively set their age to 0 and become visible in the system. This action occurs once a week during the hatching period (i.e., week 16 to 27; hatch-start and hatch-end parameters). Values chosen for these parameters were based on the hypothesis that the mean number of degree-days for brown trout to complete the egg stage was equal to 444. As we observed an average water temperature of 6_C in both streams during the spawning period, we supposed there was a delay of 10 weeks between the spawning and hatching periods.

7. Move trout of stream C downstream: Each week during the migration period (i.e., week 23 to 48; move-start and move-end parameters), each trout determines if it belongs to the group of candidate migrants. Then, if the flow in stream C is adequate, the candidate migrants that actually move to stream L become migrants.

7.1. Identify candidate migrants: Trout are candidate migrants if they are born in stream C and currently live in stream C, if their age is equal to 1 or 2 (move-min-age and move-max-age parameters), and if their length follows a normal distribution of mean and standard deviation equal to move-mean-length and move-sd-length. The status of trout meeting these criteria is set to 'candidate-migrant'.

7.2. Move candidate migrants: Young individuals often leave their home tributary to large rivers for feeding. It appears there is an apparent advantage in form of increased growth, size and thereby reproductive potential by moving instead of remaining resident in the nursery area. Several factors can trigger these movements, such as the physical state of the individuals and hydrological criteria. This migration can also be explained by density-dependent mechanisms. Juveniles that are unable to establish a territory in the nursery brook may be displaced by competition with dominant individuals. In the model, the downstream movement of candidate migrants occurs when water temperature and flow rate in stream C follow distributions of means and standard deviations equal to move-mean-temperature, move-mean-flow, move-sd-temperature and move-sd-flow, respectively. On weeks meeting these criteria, a random number between 0 and 1 is drawn for each candidate migrant. If this number is lower than the probability of moving downstream, then the current location of the candidate is changed from 'C' to 'L' and its status is set to 'migrant'. The migration of individuals is age-specific and follows a density-based logistic function given by: P = exp (A * X + B) / (1 + exp (A * X + B)), where A and B correspond to parameters move-ageZ-varA and move-ageZ-varB, respectively. Their values were obtained by calibration.

8. Move upswimming spawners back to stream L: One week after the end of the spawning period (i.e., week 18; spawn-end + 1), candidate and effective spawners that have moved to spawn return to their original stream (post-spawning homing behaviour). Their current location is changed from 'C' to 'L'.

9. Remove young trout of stream L from the system: The last week of the year (i.e., week 52), trout are randomly selected among (i) migrants from stream C currently living in stream L, with a proportion of 57% corresponding to the leaving-propC parameter (calibrated value), (ii) juveniles born in stream L and currently living in stream L, with a proportion equal to 7% (leaving-propL parameter; calibrated value). These selected trout are assumed to leave the system to search for more suitable locations, owing to their territorial behaviour.

HOW TO USE IT

Click the 'Setup' button to initialise the model. The values of all variables are imported from the file ''DemGenTrout-world.csv''. This solution was chosen because the model was relatively slow to create all the turtles (see DemGenTrout 1.0).

You can change the values of parameters with the sliders, then click the 'change parameters' button, or you can keep the default values (to restore them, click the 'restore parameters' button). Not all parameters can be changed, but the ones that were identified as the most influential in the model are included (Frank & Baret, 2013). In particular, the survival rate of fry in the brook (C-age0-survival) and the mean of the spawner condition factor (spawn-mean-cond) showed the strongest influence, followed by the survival rates of age 1 and age 2 trout in stream C (C-age1-survival and C-age2-survival), the survival rate of fry in stream L (L-age0-survival), and the von Bertalanffy growth coefficient for trout in stream L (streamL-parK).

Click the 'go' button to start the simulation (click 'go' again to stop it). The model can also be run weekly ('step-W' button), monthly ('step-M') or yearly ('step-Y'). The two plots on the left vary with time, and show the age distribution of trout in both streams (numbers of trout in each age-class are also displayed below each plot). The first plot on the right shows trout abundance in both streams, and is updated each year at week 1. The second plot represents the number of spawners moving from stream L to stream C for reproduction in winter (upswimming spawners), then moving back to stream L (downswimming spawners). For upswimming spawners, the percentage of spawners born in stream C is indicated (natal homing behaviour). The number of offspring produced in each stream is also monitored. The last plot shows the number of age-1 and age-2 trout migrating from stream C in spring and summer.

THINGS TO NOTICE AND TO TRY

As genetic outputs are derived from analyses of trout genotypes with R, only demographic outputs can be watched here.

Notice how the trout age distributions are different. Trout of age class 0 and 1 are less represented in the main river than in the Chicheron Brook, showing the nursery role of the tributary. Abundance is more stable in the main river than in the brook. You can modify the demographic structure by moving the slider corresponding to the survival rate of fry in the brook (C-age0-survival). A value close to 0 will induce trout extinction in both streams, a value close to 1 will cause the reverse effect (i.e., demographic explosion).

Look at the annual number of spawners ascending or descending the tributary: the ratio is around 50%. You can alter the mortality of spawners during their stay in the brook by moving the slider corresponding to the predation-factor parameter. The number of upswimming spawners can also be changed by varying the moved-prop parameter.

EXTENDING THE MODEL

Possibilities for improving the realism of the DemGenTrout model are (i) the consideration of density-dependence in other submodels than the downstream movement of juveniles (e.g., survival and growth processes), (ii) the integration of the quantitative genetics dimension, (iii) the inclusion of additional environmental factors such as rainfall.

The extension of the model spatial scale could also be envisaged: we could switch from the 6 km_ system comprising the Lesse River / Chicheron Brook, to a 1343 km_ system formed by the Lesse watershed. Several populations would interact, and the impact of brown trout behaviour on the exchange of genetic material among populations could therefore be studied.

RELATED MODELS

DemGenTrout 1.0: This is the first version of the DemGenTrout model, as described in Frank & Baret (in press).

DemGenTrout 1.2: This version includes features of DemGenTrout 1.1 (faster initialisation) and simulates a barrier to upstream spawning migration.

DemGenTrout 1.3: This version includes features of DemGenTrout 1.1 (faster initialisation) and simulates stocking with hatchery trout.

CREDITS AND CITATION

The development of the model was funded by the "Fonds pour la formation a la Recherche dans l'Industrie et dans l'Agriculture" (F.R.I.A.). During the formulation phase, the report describing the inSTREAM model of Railsback et al. (2009) has been an inspirational source. For the programming phase, many of the sample NetLogo models were of great use.

To refer to this model in academic publications, please use: Frank, B.M., Baret, P.V. (2013). Simulating brown trout demogenetics in a river/nursery brook system: The individual-based model DemGenTrout. Ecological Modelling 248: 184-202.

COPYRIGHT AND LICENSE

DemGenTrout 1.1 (2012)

Beatrice M. Frank

Earth and Life Institute

Universite catholique de Louvain

Beatrice.Frank@gmail.com

CC BY-NC-SA 3.0

This work is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/.

Comments and Questions

About DemGenTrout 1.1

In comparison with DemGenTrout 1.0, the initialisation of this version is faster. An applet of the model is available at: http://sites-final.uclouvain.be/gena-truites/DemGenTrout/DemGenTrout-V1-1.html

Posted over 6 years ago

Click to Run Model

;; DemGenTrout 1.1 (2012)
;;
;; Developed by
;; Beatrice M. Frank
;; Earth and Life Institute
;; Universite catholique de Louvain
;; Beatrice.Frank@gmail.com
;; 
;; Last updated: September 2012

breed[ trout ]
breed[ dead ]
breed[ gone ]

;; trout state variables
trout-own[ sex genotype natal-stream week-of-birth week-of-birth2 ;; fixed
  current-stream age stage status body-length body-weight condition-factor num-offspring
  spawned? moved-to-spawn? returned? ]

;; gone trout state variables
gone-own[ sex genotype natal-stream week-of-birth week-of-birth2 ;; fixed
  current-stream age stage status body-length body-weight condition-factor num-offspring
  spawned? moved-to-spawn? returned? ]

;; dead trout state variables
dead-own[ sex genotype natal-stream week-of-birth week-of-birth2 ;; fixed
  current-stream age stage status body-length body-weight condition-factor num-offspring
  spawned? moved-to-spawn? returned? ]
  
;; stream state variables
patches-own [ stream flow temperature ] 

globals[ week week-of-year month-of-year year
 ;; trout initialisation
 prop-C prop-L init-N
 prop-age0-C prop-age1-C prop-age2-C prop-age3-C
 prop-age0-L prop-age1-L prop-age2-L
 meanl-age0-C meanl-age0-L meanl-age1-C meanl-age1-L meanl-age2-C meanl-age2-L meanl-age3-C meanl-age3-L
 sdl-age0-C sdl-age0-L sdl-age1-C sdl-age1-L sdl-age2-C sdl-age2-L sdl-age3-C sdl-age3-L 
 trdown-age0-C trdown-age0-L trdown-age1-C trdown-age1-L trdown-age2-C trdown-age2-L trdown-age3-C trdown-age3-L
 trup-age0-C trup-age0-L trup-age1-C trup-age1-L trup-age2-C trup-age2-L trup-age3-C trup-age3-L
 meanl-birth-C sdl-birth-C meanl-birth-L sdl-birth-L trdown-C trup-C trdown-L trup-L 
 mean-bcf-C sd-bcf-C mean-bcf-L sd-bcf-L
 propC-inL 
 ;; stream input data
 streamC-discharge streamL-discharge streamC-temperature streamL-temperature 
 mean-dischargeC sd-dischargeC mean-dischargeL sd-dischargeL
 mean-temperatureC sd-temperatureC mean-temperatureL sd-temperatureL
 max-dischargeC max-dischargeL

 ;; survival parameters 
 streamC-age0-survival streamC-age1-survival streamC-age2-survival streamC-age3-survival
 streamL-age0-survival streamL-age1-survival streamL-age2-survival streamL-age3-survival 
 streamC-age0-survival-list streamC-age1-survival-list streamC-age2-survival-list streamC-age3-survival-list
 streamL-age0-survival-list streamL-age1-survival-list streamL-age2-survival-list streamL-age3-survival-list  
; predation-factor
 ;; growth parameters
; streamC-parK streamL-parK
 streamC-max-length streamC-parA streamC-parB streamL-max-length streamL-parA streamL-parB   
 ;; spawning parameters
 ;spawn-start spawn-end moved-prop spawn-mean-length spawn-mean-cond
 spawn-sd-length spawn-sd-cond spawn-min-age
 spawn-mean-temperature spawn-sd-temperature spawn-mean-flow spawn-sd-flow
 streamC-capacity offprod offprod-min-length offprod-max-length offprod-min
 offspringC-varA offspringC-varB offspringL-varA offspringL-varB
 ;offprodC-max offprodL-max 
 length-heritability 
 ;; hatching parameters
 hatch-start hatch-end
 ;; downstream movement parameters
 ;move-start move-end move-min-age move-max-age
 move-mean-length move-sd-length 
 move-age1-prob move-age1-varB move-age1-varA
 move-age2-prob move-age2-varB move-age2-varA 
 move-mean-temperature move-sd-temperature move-mean-flow move-sd-flow
 ;; leaving stream L forever parameters
 leaving-propC leaving-propL
 
 out-year out-month out-week
 ]  

to setup 
  clear-all      
  no-display 
  ;; controlling the random numbers
  random-seed 1223251200
  
  ;; first week of the year = 1st to 7th October included
  set week 1
  set week-of-year 1
  set month-of-year 1
  set year 1 
  
  import-world "DemGenTrout-world.csv"
  reset-ticks
end 

to restore-defaults
  ;; restore the default parameters         
  ;;; input parameters ;;;
  file-open "init-params.txt"
  ;; weekly survival rates for trout in stream C
  if file-read = "streamC.age0.survival" [set streamC-age0-survival file-read]
  if file-read = "streamC.age1.survival" [set streamC-age1-survival file-read] 
  if file-read = "streamC.age2.survival" [set streamC-age2-survival file-read]
  if file-read = "streamC.age3.survival" [set streamC-age3-survival file-read]
  ;; weekly survival rates for trout in stream L  
  if file-read = "streamL.age0.survival" [set streamL-age0-survival file-read]  
  if file-read = "streamL.age1.survival" [set streamL-age1-survival file-read]
  if file-read = "streamL.age2.survival" [set streamL-age2-survival file-read]
  if file-read = "streamL.age3.survival" [set streamL-age3-survival file-read]  
  if file-read = "predation.factor" [set predation-factor file-read]   
  ;; growth parameters
  if file-read = "streamC.parK" [set streamC-parK file-read]
  if file-read = "streamL.parK" [set streamL-parK file-read]  
  if file-read = "streamC.max.length" [set streamC-max-length file-read]
  if file-read = "streamL.max.length" [set streamL-max-length file-read]  
  if file-read = "streamC.parA" [set streamC-parA file-read]
  if file-read = "streamL.parA" [set streamL-parA file-read]  
  if file-read = "streamC.parB" [set streamC-parB file-read]
  if file-read = "streamL.parB" [set streamL-parB file-read]           
  ;; spawning parameters
  if file-read = "spawn.start" [set spawn-start file-read]
  if file-read = "spawn.end" [set spawn-end file-read] 
  if file-read = "spawn.min.age" [set spawn-min-age file-read]      
  if file-read = "spawn.mean.length" [set spawn-mean-length file-read]  
  if file-read = "spawn.sd.length" [set spawn-sd-length file-read]   
  if file-read = "spawn.mean.cond" [set spawn-mean-cond file-read] 
  if file-read = "spawn.sd.cond" [set spawn-sd-cond file-read]  
  if file-read = "moved.prop" [set moved-prop file-read]       
  if file-read = "spawn.mean.flow" [set spawn-mean-flow file-read]  
  if file-read = "spawn.sd.flow" [set spawn-sd-flow file-read]   
  if file-read = "streamC.capacity" [set streamC-capacity file-read]    
  if file-read = "offprod.min.length" [set offprod-min-length file-read] 
  if file-read = "offprod.max.length" [set offprod-max-length file-read]     
  if file-read = "offprod.min" [set offprod-min file-read]         
  if file-read = "offprodC.max" [set offprodC-max file-read] 
  if file-read = "offprodL.max" [set offprodL-max file-read] 
  if file-read = "length.heritability" [set length-heritability file-read]     
  ;; hatching parameters
  if file-read = "hatch.start" [set hatch-start file-read]
  if file-read = "hatch.end" [set hatch-end file-read]      
  ;; downstream movement parameters
  if file-read = "move.start" [set move-start file-read] 
  if file-read = "move.end" [set move-end file-read] 
  if file-read = "move.min.age" [set move-min-age file-read] 
  if file-read = "move.max.age" [set move-max-age file-read]  
  if file-read = "move.mean.length" [set move-mean-length file-read] 
  if file-read = "move.sd.length" [set move-sd-length file-read]     
  if file-read = "move.age1.varA" [set move-age1-varA file-read] 
  if file-read = "move.age1.varB" [set move-age1-varB file-read]  
  if file-read = "move.age2.varA" [set move-age2-varA file-read] 
  if file-read = "move.age2.varB" [set move-age2-varB file-read] 
  if file-read = "move.mean.temperature" [set move-mean-temperature file-read] 
  if file-read = "move.sd.temperature" [set move-sd-temperature file-read] 
  if file-read = "move.mean.flow" [set move-mean-flow file-read] 
  if file-read = "move.sd.flow" [set move-sd-flow file-read]            
  ;; leaving streamL forever parameters
  if file-read = "leaving.propC" [set leaving-propC file-read]  
  if file-read = "leaving.propL" [set leaving-propL file-read]    
  file-close

  set C-age0-survival 0.39 set C-age1-survival 0.37 set C-age2-survival 0.85 set C-age3-survival 0.70
  set L-age0-survival 0.38 set L-age1-survival 0.74 set L-age2-survival 0.87 set L-age3-survival 0.69
end     

to set-parameters 
  ;; weekly survival rates for trout in stream C  
  file-open "streamC-age0-survival.txt"  
  set streamC-age0-survival-list file-read file-close
  file-open "streamC-age1-survival.txt"  
  set streamC-age1-survival-list file-read file-close  
  file-open "streamC-age2-survival.txt"  
  set streamC-age2-survival-list file-read file-close
  file-open "streamC-age3-survival.txt"  
  set streamC-age3-survival-list file-read file-close  
  ;; weekly survival rates for trout in stream L  
  file-open "streamL-age0-survival.txt"  
  set streamL-age0-survival-list file-read file-close  
  file-open "streamL-age1-survival.txt"  
  set streamL-age1-survival-list file-read file-close 
  file-open "streamL-age2-survival.txt"  
  set streamL-age2-survival-list file-read file-close     
  file-open "streamL-age3-survival.txt"  
  set streamL-age3-survival-list file-read file-close   
  
  let seq (list 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
                     0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19
                     0.20 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29
                     0.30 0.31 0.32 0.33 0.34 0.35 0.36 0.37 0.38 0.39
                     0.40 0.41 0.42 0.43 0.44 0.45 0.46 0.47 0.48 0.49
                     0.50 0.51 0.52 0.53 0.54 0.55 0.56 0.57 0.58 0.59
                     0.60 0.61 0.62 0.63 0.64 0.65 0.66 0.67 0.68 0.69
                     0.70 0.71 0.72 0.73 0.74 0.75 0.76 0.77 0.78 0.79
                     0.80 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89
                     0.90 0.91 0.92 0.93 0.94 0.95 0.96 0.97 0.98 0.99 1.00)
  let index (list 0  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19
                 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39
                 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59
                 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
                 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99
                 100)
  
  (foreach seq index
    [
      if C-age0-survival = ?1 [set streamC-age0-survival item ?2 streamC-age0-survival-list]
      if C-age1-survival = ?1 [set streamC-age1-survival item ?2 streamC-age1-survival-list]
      if C-age2-survival = ?1 [set streamC-age2-survival item ?2 streamC-age2-survival-list]
      if C-age3-survival = ?1 [set streamC-age3-survival item ?2 streamC-age3-survival-list]                  
      if L-age0-survival = ?1 [set streamL-age0-survival item ?2 streamL-age0-survival-list]
      if L-age1-survival = ?1 [set streamL-age1-survival item ?2 streamL-age1-survival-list]
      if L-age2-survival = ?1 [set streamL-age2-survival item ?2 streamL-age2-survival-list]
      if L-age3-survival = ?1 [set streamL-age3-survival item ?2 streamL-age3-survival-list]
    ])    
  
  set offprod offprodC-max
end 

;;;;;;;;;;;;;;;;;;;;;;;;;;;;; SCHEDULING ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

to go 
  no-display   
  tick 
  if ticks = 1820 [ user-message "Input data are only available for 35 years."  stop ]
  ;; a year = 52 weeks  
  set week-of-year (week mod 52) + 1
  set month-of-year (int(week-of-year / 4.5) mod 52) + 1 
  set year (int(week / 52)) + 1 ; mod 52  
  
  ;; each week

;-- 1. Update stream hydrological conditions ----------------------------------------------------------
  
  ask patches [ update-stream-conditions ] 
  
;-- 2. Kill trout in each stream ----------------------------------------------------------------------

  ask trout [ trout-die ]      
    
;-- 3. Update trout length, weight and condition factor -----------------------------------------------
  
  ask trout [ trout-grow ]  
      
;-- 4. Reproduce trout in each stream ----------------------------------------------------------------- 

  ;; spawning time window: duration = 11 weeks
  if week-of-year >= spawn-start and week-of-year <= spawn-end
  [   
    ;--- 4.1. Identify candidate spawners ---;
    ask trout [ trout-become-candidate-spawners ] 
    ;; select some of the candidate spawners born in stream L in order to make them reproduce in C
    let candidate-spawners-streamL trout with [ status = "candidate-spawner" and current-stream = "L" 
      and natal-stream = "L" and not moved-to-spawn? ]
    while [ (count trout with [status = "selected-candidate-spawner"]) < (moved-prop * count candidate-spawners-streamL) ]
    [ ask one-of candidate-spawners-streamL [ set status "selected-candidate-spawner" ] ]      
    
    ;--- 4.2. Move candidate spawners upstream ---;   
    ask trout [ trout-move-to-spawn ]         
    
    ;; adapt offspring production if abundance in C is strictly higher than streamC-capacity
    ifelse count trout with [ current-stream = "C" ] > streamC-capacity 
    [ set offprodC-max offprod / 10 ]
    [ set offprodC-max offprod ]        

    ;--- 4.3. Produce offspring in each stream ---;  
    trout-spawn-in-streamC
    trout-spawn-in-streamL
  ]
    
;-- 5. Increment trout age and update stage -----------------------------------------------------------

  if week-of-year = (hatch-start - 1) ;; one week before hatching
  [ 
    ask trout [ trout-update-age ]
    ask trout with [age > 6] [die!] ;; trout die at age 7    
  ]     
   
;-- 6. Reveal offspring -------------------------------------------------------------------------------   
  
  ;; hatching time window: duration = 11 weeks
  if week-of-year >= hatch-start and week-of-year <= hatch-end
  [ (foreach sort trout with [age = -1]
      [ ask ?
        [
          if (week-of-year - week-of-birth) = (hatch-start - spawn-start) ;; delay = 10 weeks
          [ set hidden? false set age 0 ]
        ]])
  ]
  
;-- 7. Move trout of stream C downstream ----------------------------------------------------------
    
  ;; migration time window: duration = 25 weeks
  if week-of-year >= move-start and week-of-year <= move-end
  [             
    ;--- 7.1. Identify candidate migrants among juveniles ---; 
    ask trout [ trout-become-candidate-migrants ]          
     
    ;--- 7.2. Update movement probabilities ---;     
    update-move-age1-prob   
    update-move-age2-prob      
            
    ;--- 7.3. Move candidate migrants ---;
    ask trout [ trout-move-downstream ]                  
  ]       

;-- 8. Move upswimming spawners back to stream L ----------------------------------------------------- 

  if week-of-year = (spawn-end + 1) ;; the week after the spawning process
  [ 
    ask trout [ trout-move-back ]
  ]                                           
   
;-- 9. Remove young trout of stream L from the system ------------------------------------------------   
  
  ;; at the end of the year
  if week-of-year = 52
  [ 
    ;; some juveniles born in C and living in stream L do not settle and go elsewhere 
    let bornC-in-streamL trout with [current-stream = "L" and natal-stream = "C" and status = "migrant"]    
    let bornC-leaving (leaving-propC * count bornC-in-streamL)
    ask n-of bornC-leaving bornC-in-streamL [ leave-streamL ]  
    ;; some juveniles born in L and living in stream L do not settle and go elsewhere  
    let bornL-in-streamL trout with [current-stream = "L" and natal-stream = "L" and (age = 1 or age = 2)]
    let bornL-leaving (leaving-propL * count bornL-in-streamL)
    ask n-of bornL-leaving bornL-in-streamL [ leave-streamL ]   
  ]                
  
; Trout reset --------------------------------------------------------------------------------------
  
  ;; at the end of the year
  if week-of-year = 52
  [        
    ;; reset trout status and boolean state variables    
    ask trout [ set status "non-spawner" set moved-to-spawn? false set spawned? false set returned? false ] 
    ;; kill all hidden turtles  
    ask gone [die]     
    ask dead [die]          
  ]

; Increment the week -------------------------------------------------------------------------------
   
  set week week + 1
end 

;;;;;;;;;;;;;;;;;;;;;;;;;;; STREAM PROCEDURE ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

;; update water temperature and flow of each stream

to update-stream-conditions
  ifelse ticks < (6 * 52)
  ;; first 6 years (2004 to 2009)
  [
    ifelse stream = "C"
    [
      set flow item week streamC-discharge
      set temperature item week streamC-temperature 
    ]
    [
      set flow item week streamL-discharge
      set temperature item week streamL-temperature
    ]  
  ]
  ;; next years (2010 to 2038)
  [
    ifelse stream = "C"
    [
      set flow min (list max-dischargeC (item week streamC-discharge + random-lognormal mean-dischargeC sd-dischargeC))
      set temperature (item week streamC-temperature + random-normal mean-temperatureC sd-temperatureC) 
    ]
    [
      set flow min (list max-dischargeL (item week streamL-discharge + random-lognormal mean-dischargeL sd-dischargeL))
      set temperature (item week streamL-temperature + random-normal mean-temperatureL sd-temperatureL)
    ]  
  ]
end 

;;;;;;;;;;;;;;;;;;;;;;;;;; TROUT PROCEDURES ;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

;; trout mortality

to trout-die
  ;; in stream C
  if stage = "fry" and current-stream = "C" [if random-float 1 > ((item week-of-year streamC-age0-survival) ^ (1 / 52)) [die!]]
  if stage = "juvenile" and age = 1 and current-stream = "C" [if random-float 1 > ((item week-of-year streamC-age1-survival) ^ (1 / 52)) [die!]]
  if stage = "juvenile" and age = 2 and current-stream = "C" [if random-float 1 > ((item week-of-year streamC-age2-survival) ^ (1 / 52)) [die!]]
  if stage = "adult" and age < 6 and current-stream = "C" [if random-float 1 > ((item week-of-year streamC-age3-survival) ^ (1 / 52)) [die!]] 
  if week-of-year >= spawn-start and week-of-year <= spawn-end and moved-to-spawn? = true and current-stream = "C" 
   [if random-float 1 > (((item week-of-year streamC-age3-survival) / predation-factor) ^ (1 / (spawn-end - spawn-start))) [die!]] 
  ;; in stream L
  if stage = "fry" and current-stream = "L" [if random-float 1 > ((item week-of-year streamL-age0-survival) ^ (1 / 52)) [die!]]
  if stage = "juvenile" and age = 1 and current-stream = "L" [if random-float 1 > ((item week-of-year streamL-age1-survival) ^ (1 / 52)) [die!]]
  if stage = "juvenile" and age = 2 and current-stream = "L" [if random-float 1 > ((item week-of-year streamL-age2-survival) ^ (1 / 52)) [die!]]
  if stage = "adult" and age < 6 and current-stream = "L" [if random-float 1 > ((item week-of-year streamL-age3-survival) ^ (1 / 52)) [die!]]
end 

;; individual growth

to trout-grow
  ifelse current-stream = "C"
  [
    ;; von Bertalanffy growth equation for trout in stream C
    set body-length (body-length + (streamC-parK * (streamC-max-length - body-length)))
    let healthy-weight (streamC-parA * ((body-length) ^ streamC-parB)) ;; weight-length regression 
    set body-weight (condition-factor * healthy-weight)      
  ]
  [
    ;; von Bertalanffy growth equation for trout in stream L
    set body-length (body-length + (streamL-parK * (streamL-max-length - body-length)))  
    let healthy-weight (streamL-parA * ((body-length) ^ streamL-parB)) ;; weight-length regression
    set body-weight (condition-factor * healthy-weight)      
  ]
  set condition-factor max (list 0.5 (condition-factor + (- 0.008 + random-float 0.016)))
  ;; test for negative length and negative weight
  if body-length < 0 [ user-message "Negative fish length" ]
  if body-weight < 0 [ user-message "Negative fish weight" ]    
end 

to die!
  set breed dead
  hide-turtle
end 

;; identification of candidate spawners

to trout-become-candidate-spawners
  if age >= spawn-min-age and status != "spawner" and not moved-to-spawn? 
  [ if random-normal spawn-mean-length spawn-sd-length < body-length and 
    random-normal spawn-mean-cond spawn-sd-cond < condition-factor
    [ set status "candidate-spawner" ] ]  
end 

;; movement to stream C for reproduction

to trout-move-to-spawn
 let temperature-streamL [temperature] of one-of patches with [ stream = "L" ]      
 let flow-streamL [flow] of one-of patches with [ stream = "L" ] 
 ;; spawners born in C or in L move to C
 if (status = "candidate-spawner" and not moved-to-spawn? and current-stream = "L" and natal-stream = "C") 
   or (status = "selected-candidate-spawner" and not moved-to-spawn?)     
 [ if random-lognormal spawn-mean-flow spawn-sd-flow < flow-streamL   
   [ set current-stream "C" set xcor 1 set moved-to-spawn? true set status "candidate-spawner" ] ]  
end 

;; update trout age and stage

to trout-update-age
  set age age + 1
  trout-update-stage 
end   

;; update trout stage: fry, juvenile, adult 

to trout-update-stage
  if age < 1 [ set stage "fry" stop ]
  if age >= 1 and age <= 2 [ set stage "juvenile"stop ]
  if age > 2 [ set stage "adult" stop ] 
end 

;; reproduction

to trout-spawn-in-streamC
  let spawners trout with [ status = "candidate-spawner" or status = "spawner" and current-stream = "C" ]  
  ifelse count spawners < 2
  [ stop ]
  [  
    ;; length of parents when born (VBLF backwards)
    let born-length-parents VBLF-reverse spawners    
    ;; mean and variance of length of parents when born
    if length born-length-parents < 2 [ stop ]
    let mean-parents mean born-length-parents
    let var-parents variance born-length-parents
    let num-crossover (floor (count spawners / 2))  
    ;; polygamous mating with satellite males (not in good condition)    
    repeat num-crossover
    [
      let num-males 1 + random 4 ; number of males drawn randomly from an uniform distribution from 1 to 4
      let all-parent-M spawners with [ sex = "M" ]
      if count all-parent-M < num-males [ stop ]         
      let no-parent-M max-n-of num-males spawners with [ sex = "M" ] [condition-factor]
      let sel-parent-M all-parent-M with [ not member? self no-parent-M and not spawned? ]  
      if count spawners with [ sex = "F" and not spawned? ] <= 1 [ stop ] 
      if count sel-parent-M < num-males [ stop ]  
      ;; select the parents: first, the spawners that have moved to spawn, then the other spawners  
      let parent-M n-of num-males sel-parent-M
      let parent-F max-one-of spawners with [ sex = "F" and not spawned? ] [condition-factor]     
      if (count spawners with [ sex = "M" and not spawned? and moved-to-spawn? ] - num-males) >= num-males  
      [ 
        if any? spawners with [ sex = "M" and not spawned? and moved-to-spawn? ] 
        [
          set all-parent-M spawners with [ sex = "M" and moved-to-spawn? ]
          set no-parent-M max-n-of num-males spawners with [ sex = "M" and moved-to-spawn? ] [condition-factor]
          set sel-parent-M all-parent-M with [ not member? self no-parent-M and not spawned? ]           
          set parent-M n-of num-males sel-parent-M
        ]
      ]
      if any? spawners with [ sex = "F" and not spawned? and moved-to-spawn?]
      [ set parent-F max-one-of spawners with [ sex = "F" and not spawned? and moved-to-spawn?] [condition-factor] ]
      ;; length deviances of parents M and F from mean-parents
      let dummy VBLF-reverse parent-M
      let dev-M 0
      let temp (random-normal meanl-birth-C sdl-birth-C)
      if temp < trdown-C [set temp trdown-C]   
      if temp > trup-C [set temp trup-C]                
      if length dummy = 0 [ set dev-M (mean-parents - temp) ]      
      if length dummy = 1 [ set dev-M (mean-parents - (item 0 dummy)) ]        
      if length dummy > 1 [ set dev-M (mean-parents - mean dummy) ]
      let dev-F (mean-parents - (VBLF-reverse parent-F)) 
      ask parent-F
      [
        set spawned? true
        set status "spawner"
        set xcor 1
        let num-fry update-num-offspringC [body-length] of parent-F   
        hatch-trout num-fry
        [
          set hidden? true ;; hide trout until hatching
          set age -1 set stage "fry"  
          ; genotype of parent-M is a mixture of the several genotypes of males
          let genotype-mix mixture ([genotype] of parent-M)
          ;; define genotypes for eggs
          let egg-genotype cross-genotype (genotype-mix) ([genotype] of parent-F)                
          set genotype egg-genotype                         
          set xcor 1
          set current-stream "C"
          set natal-stream "C"
          set num-offspring 0
          set week-of-birth week-of-year      
          set week-of-birth2 (week + 1)          
          ifelse random-float 1 > 0.5 [ set sex "M" ] [ set sex "F" ]
          set status "non-spawner" set spawned? false
          set moved-to-spawn? false set returned? false          
          ;; set body-length as function of parents' length and environment
          let genetic-dev (sqrt length-heritability) * (dev-M + dev-F)    
          let envi-var ((1 - length-heritability) ^ 2) * var-parents
          let envi-dev random-normal 0 (sqrt envi-var) 
          set body-length mean-parents + genetic-dev + envi-dev
          let temp2 (random-normal meanl-birth-C sdl-birth-C)  
          if temp2 < trdown-C [set temp2 trdown-C]   
          if temp2 > trup-C [set temp2 trup-C]                
          if body-length < trdown-C or body-length > trup-C [ set body-length temp2 ]           
          ;; set body-weight and condition-factor
          set condition-factor (random-normal mean-bcf-C sd-bcf-C)         
          set body-weight (condition-factor * ((body-length) ^ 3) / 100000)       
        ] 
        set num-offspring num-fry
      ]
      ask parent-M
      [
        ;set spawned? true ;; shut because males can reproduce more than once
        set status "spawner" 
        set xcor 1        
        set num-offspring [num-offspring] of parent-F               
      ] 
    ]
  ]               
end  

to trout-spawn-in-streamL
  let spawners trout with [ status = "candidate-spawner" or status = "spawner" and current-stream = "L" ]
  ifelse count spawners < 2
  [ stop ]
  [  
    let born-length-parents VBLF-reverse spawners 
    ;; mean and variance of length of parents when born
    if length born-length-parents < 2 [ stop ]  
    let mean-parents mean born-length-parents
    let var-parents variance born-length-parents  
    let num-crossover (floor (count spawners / 2))   
    ;; polygamous mating with satellite males (not in good condition)    
    repeat num-crossover  
    [
      let num-males 1 + random 4 ; number of males drawn randomly from an uniform distribution from 1 to 4
      let all-parent-M spawners with [ sex = "M" ]
      if count all-parent-M < num-males [ stop ]         
      let no-parent-M max-n-of num-males spawners with [ sex = "M" ] [condition-factor]
      let sel-parent-M all-parent-M with [ not member? self no-parent-M and not spawned? ]  
      if count spawners with [ sex = "F" and not spawned? ] <= 1 [ stop ] 
      if count sel-parent-M < num-males [ stop ]  
      ;; select the parents: first, the spawners that have moved to spawn, then the other spawners  
      let parent-M n-of num-males sel-parent-M
      let parent-F max-one-of spawners with [ sex = "F" and not spawned? ] [condition-factor]     
      if (count spawners with [ sex = "M" and not spawned? and moved-to-spawn? ] - num-males) >= num-males  
      [ 
        if any? spawners with [ sex = "M" and not spawned? and moved-to-spawn? ] 
        [
          set all-parent-M spawners with [ sex = "M" and moved-to-spawn? ]
          set no-parent-M max-n-of num-males spawners with [ sex = "M" and moved-to-spawn? ] [condition-factor]
          set sel-parent-M all-parent-M with [ not member? self no-parent-M and not spawned? ]           
          set parent-M n-of num-males sel-parent-M
        ]
      ]
      if any? spawners with [ sex = "F" and not spawned? and moved-to-spawn?]
      [ set parent-F max-one-of spawners with [ sex = "F" and not spawned? and moved-to-spawn?] [condition-factor] ]
      ;; length deviances of parents M and F from mean-parents
      let dummy VBLF-reverse parent-M
      let dev-M 0
      let temp (random-normal meanl-birth-L sdl-birth-L)
      if temp < trdown-L [set temp trdown-L]   
      if temp > trup-L [set temp trup-L]                
      if length dummy = 0 [ set dev-M (mean-parents - temp) ]            
      if length dummy = 1 [ set dev-M (mean-parents - (item 0 dummy)) ]        
      if length dummy > 1 [ set dev-M (mean-parents - mean dummy) ]    
      let dev-F (mean-parents - (VBLF-reverse parent-F))      
      ask parent-F
      [
        set spawned? true
        set status "spawner"
        set xcor 0
        let num-fry update-num-offspringL [body-length] of parent-F 
        hatch-trout num-fry
        [
          set hidden? true ;; hide trout until hatching
          set age -1 set stage "fry"
          ; genotype of parent-M is a mixture of the several genotypes of males
          let genotype-mix mixture ([genotype] of parent-M)
          ;; define genotypes for eggs
          let egg-genotype cross-genotype (genotype-mix) ([genotype] of parent-F)              
          set genotype egg-genotype                                 
          set xcor 0
          set current-stream "L"
          set natal-stream "L"  
          set num-offspring 0   
          set week-of-birth week-of-year      
          set week-of-birth2 (week + 1)              
          ifelse random-float 1 > 0.5 [ set sex "M" ] [ set sex "F" ]    
          set status "non-spawner" set spawned? false
          set moved-to-spawn? false set returned? false                   
          ;; set body-length as function of parents' length and environment
          let genetic-dev (sqrt length-heritability) * (dev-M + dev-F)     
          let envi-var ((1 - length-heritability) ^ 2) * var-parents
          let envi-dev random-normal 0 (sqrt envi-var) 
          set body-length mean-parents + genetic-dev + envi-dev
          let temp2 (random-normal meanl-birth-L sdl-birth-L)  
          if temp2 < trdown-L [set temp2 trdown-L]   
          if temp2 > trup-L [set temp2 trup-L]                         
          if body-length < trdown-L or body-length > trup-L [ set body-length temp2 ]       
          ;; set body-weight and condition-factor
          set condition-factor (random-normal mean-bcf-L sd-bcf-L)
          set body-weight (condition-factor * ((body-length) ^ 3) / 100000)              
        ]
        set num-offspring num-fry
      ]
      ask parent-M
      [
        ;set spawned? true ;; shut because males can reproduce more than once
        set status "spawner"
        set xcor 0      
        set num-offspring [num-offspring] of parent-F
      ]       
    ]  
  ]      
end   

;; stream C

to-report update-num-offspringC [ female-length ] ;; linear function
  ;; compute the intermediate variables for the linear offspring production function for trout of stream C
  set offspringC-varA (offprodC-max - offprod-min) / (offprod-max-length - offprod-min-length)
  set offspringC-varB offprod-min - (offspringC-varA * offprod-min-length) 
  report ((offspringC-varA * female-length) + offspringC-varB)
end 

;; stream L

to-report update-num-offspringL [ female-length ] ;; linear function
  ;; compute the intermediate variables for the linear offspring production function  for trout of stream C
  set offspringL-varA (offprodL-max - offprod-min) / (offprod-max-length - offprod-min-length)
  set offspringL-varB offprod-min - (offspringL-varA * offprod-min-length)    
  report ((offspringL-varA * female-length) + offspringL-varB)
end 

to-report cross-genotype [ genotype1 genotype2 ]
  let new-genotype []
  let bit 0
  repeat length genotype1
  [
    let choice random 4
    if choice = 0
    [ set new-genotype (lput (word(substring (item bit genotype1) 0 4)(substring (item bit genotype2) 0 3)) new-genotype)]
    if choice = 1
    [ set new-genotype (lput (word(substring (item bit genotype1) 0 3) (substring (item bit genotype2) 3 7)) new-genotype)]
    if choice = 2
    [ set new-genotype (lput (word(substring (item bit genotype2) 0 3) (substring (item bit genotype1) 3 7)) new-genotype)]
    if choice = 3
    [ set new-genotype (lput (word(substring (item bit genotype1) 4 7) (substring (item bit genotype2) 3 7)) new-genotype)]      
    set bit bit + 1
  ]
  report new-genotype
end 

to-report mixture [ genotype-temp ]
  let genotype-mix []
  let bit 0
  repeat length one-of genotype-temp
  [
    let choice random (length genotype-temp) ; choose randomly one genotype among n males
    set genotype-mix lput item bit (item choice genotype-temp) genotype-mix
    set bit bit + 1
  ] 
  report genotype-mix
end 

to-report VBLF-reverse [ trout-indiv ]  
  ifelse is-agentset? trout-indiv
  [
    let length-when-born []
    (foreach sort trout-indiv
      [
        let time 0
        let length-diff 0
        let length-when-born-temp 0
        ask ?
        [
          set time (week - week-of-birth2)
          ifelse natal-stream = "C" 
          [ 
            set length-diff (streamC-max-length - body-length)
            set length-when-born-temp (streamC-max-length - (length-diff / ((1 - streamC-parK) ^ (time))))            
          ]
          [ 
            set length-diff (streamL-max-length - body-length)
            set length-when-born-temp (streamL-max-length - (length-diff / ((1 - streamL-parK) ^ (time))))            
          ]
        ]     
        set length-when-born lput length-when-born-temp length-when-born                
      ])
    report length-when-born 
  ]  
  [
    let time (week - ([week-of-birth2] of trout-indiv))      
    ifelse [natal-stream] of trout-indiv = "C"
    [ 
      let length-when-born (streamC-max-length - ((streamC-max-length - [body-length] of trout-indiv) / ((1 - streamC-parK) ^ (time)))) 
      report length-when-born
    ]
    [ 
      let length-when-born (streamL-max-length - ((streamL-max-length - [body-length] of trout-indiv) / ((1 - streamL-parK) ^ (time)))) 
      report length-when-born  
    ]   
  ]   
end 

to trout-become-candidate-migrants
  if (age = move-min-age or age = move-max-age) and current-stream = "C" and natal-stream = "C" 
  [ if random-normal move-mean-length move-sd-length < body-length
    [ set status "candidate-migrant" ] ]
end 

;; movement from C to L

to trout-move-downstream
  let temperature-streamC [temperature] of one-of patches with [ stream = "C" ]  
  let flow-streamC [flow] of one-of patches with [ stream = "C" ]    
  if status = "candidate-migrant"
  [ if random-normal move-mean-temperature move-sd-temperature < temperature-streamC 
    and random-lognormal move-mean-flow move-sd-flow < flow-streamC
    [ ifelse age = 1
      ;; juveniles age 1
      [ if random-float 1 < move-age1-prob
        [ set xcor 0 set current-stream "L" set status "migrant" ] ]
      ;; juveniles age 2
      [ if random-float 1 < move-age2-prob
        [ set xcor 0 set current-stream "L" set status "migrant" ] ]
    ] 
  ]   
end     

to update-move-age1-prob ;; logistic function
  let nb1 count trout with [current-stream = "C" and body-length > 70]
  let Z1 (exp((move-age1-varA * nb1) + move-age1-varB))
  set move-age1-prob (Z1 / (1 + Z1))  
end 

to update-move-age2-prob ;; logistic function
  let nb2 count trout with [current-stream = "C" and body-length > 70]
  let Z1 (exp((move-age2-varA * nb2) + move-age2-varB))
  set move-age2-prob (Z1 / (1 + Z1))  
end 

to trout-move-back
  ;; candidate spawners return to previous stream
  if (status = "candidate-spawner" and current-stream = "C" and moved-to-spawn?)
  ;; effective spawners return to previous stream
    or (status = "spawner" and current-stream = "C" and moved-to-spawn?)
  [ set current-stream "L" set xcor 0 set returned? true ]  
end 

to leave-streamL
  set breed gone
  hide-turtle
end 

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

to-report random-lognormal [mu sigma]
  let S sqrt(ln(((sigma / mu) ^ 2) + 1))
  let M ln (mu) - ((S ^ 2) / 2)
  report exp(random-normal M S)
end 

There is only one version of this model, created over 6 years ago by Béatrice Frank.

Attached files

File Type Description Last updated
DemGenTrout-V1-1_data.zip data External files for 'DemGenTrout 1.1' about 6 years ago, by Béatrice Frank Download