predation.hpp File Reference

Go to the source code of this file.

Functions
template<class Type>
void	calculate_vulnerability (matrix< Type > &vulnerability, vector< Type > &vulnerability_other, int nspp, const vector< int > &suitMode, matrix< Type > &log_phi)
	Transforms predator-prey preference parameters into vulnerability.
template<class Type>
void	calculate_parametric_suitability (array< Type > &suitability, array< Type > &suit_other, int nspp, int nyrs, const vector< int > &nsex, const vector< int > &nages, const vector< int > &suitMode, array< Type > &length_hat, array< Type > &weight_hat, matrix< Type > &vulnerability, const vector< Type > &vulnerability_other, const vector< Type > &gam_a, const vector< Type > &gam_b)
	Parametric suitability (Gamma, Lognormal, Normal) from size ratios.
template<class Type>
void	calculate_msvpa_suitability (array< Type > &stom_div_bio, array< Type > &suitability, array< Type > &suit_other, int nspp, int nyrs, const vector< int > &suit_styr, const vector< int > &suit_endyr, const vector< int > &nyrs_suit, int msmMode, const vector< int > &nsex, const vector< int > &nages, const vector< int > &suitMode, array< Type > &diet_prop, array< Type > &avgN_at_age, array< Type > &weight_hat, array< Type > &other_food_diet_prop)
	MSVPA/empirical suitability, from stomach content and prey biomass.
template<class Type>
void	calculate_msvpa_predation (array< Type > &avail_food, array< Type > &avail_food_dB0, array< Type > &avail_food_dBF, array< Type > &M2_at_age, array< Type > &M2_at_age_dB0, array< Type > &M2_at_age_dBF, array< Type > &M2_prop, array< Type > &B_eaten, array< Type > &B_eaten_as_prey, array< Type > &diet_prop_hat, array< Type > &other_diet_prop_hat, int nspp, int nyrs, const vector< int > &nsex, const vector< int > &nages, int msmMode, array< Type > &avgN_at_age, array< Type > &avgN_at_age_dB0, array< Type > &avgN_at_age_dBF, array< Type > &suitability, array< Type > &suit_other, array< Type > &weight_hat, array< Type > &ration, const vector< Type > &other_food)
	Available food, predation mortality (M2), and consumed biomass (MSVPA).

Function Documentation

calculate_msvpa_predation()

template<class Type>

void calculate_msvpa_predation	(	array< Type > &	avail_food,
		array< Type > &	avail_food_dB0,
		array< Type > &	avail_food_dBF,
		array< Type > &	M2_at_age,
		array< Type > &	M2_at_age_dB0,
		array< Type > &	M2_at_age_dBF,
		array< Type > &	M2_prop,
		array< Type > &	B_eaten,
		array< Type > &	B_eaten_as_prey,
		array< Type > &	diet_prop_hat,
		array< Type > &	other_diet_prop_hat,
		int	nspp,
		int	nyrs,
		const vector< int > &	nsex,
		const vector< int > &	nages,
		int	msmMode,
		array< Type > &	avgN_at_age,
		array< Type > &	avgN_at_age_dB0,
		array< Type > &	avgN_at_age_dBF,
		array< Type > &	suitability,
		array< Type > &	suit_other,
		array< Type > &	weight_hat,
		array< Type > &	ration,
		const vector< Type > &	other_food )

Available food, predation mortality (M2), and consumed biomass (MSVPA).

Available food is suitability-weighted prey biomass plus "other food": prey numbers (thousands of fish) x weight (kg) = mt, and other_food is mt. Under msmMode = 2 (Type III) the prey term is squared, so the denominator is not a biomass there. M2 divides each predator's consumption (avgN x ration) by that denominator, so a LARGER available-food denominator gives LESS predation mortality. The dB0 (F = 0) and dBF (F = Ftarget) arrays repeat the calculation for the dynamic reference points. Prey biomass and the predator scalars are each computed once outside the deepest loop.

Parameters

avail_food	Available food (mt) [nspp, max_sex, max_age, nyrs], updated.
avail_food_dB0	Available food at F=0, updated.
avail_food_dBF	Available food at F=Ftarget, updated.
M2_at_age	Predation mortality (yr^-1), updated; [nspp, max_sex, max_age, nyrs].
M2_at_age_dB0	Predation mortality at F=0, updated.
M2_at_age_dBF	Predation mortality at F=Ftarget, updated.
M2_prop	A predator age's contribution to prey M2, not a share of it.
B_eaten	Prey biomass eaten by a specific predator age (mt), updated.
B_eaten_as_prey	Prey biomass eaten by all predators (mt), updated.
diet_prop_hat	Predicted diet proportions, updated.
other_diet_prop_hat	Predicted proportion of "other food" in diet, updated.
nspp	Number of species.
nyrs	Years, hindcast plus projection.
nsex	Sexes modelled per species: 1 = combined, 2 = both.
nages	Age bins per species.
msmMode	Predation mode (1 = Type II MSVPA, Holsman et al. 2015; 2 = Type III MSVPA, which squares the prey term).
avgN_at_age	Average numbers-at-age (thousands of fish).
avgN_at_age_dB0	Average numbers-at-age at F=0.
avgN_at_age_dBF	Average numbers-at-age at F=Ftarget.
suitability	Calculated suitability.
suit_other	Suitability for "other food".
weight_hat	Estimated weights-at-age (kg).
ration	Individual consumption (kg/predator/yr); consumption_at_age.
other_food	"Other food" biomass by predator (mt).

Definition at line 349 of file predation.hpp.

calculate_msvpa_suitability()

template<class Type>

void calculate_msvpa_suitability	(	array< Type > &	stom_div_bio,
		array< Type > &	suitability,
		array< Type > &	suit_other,
		int	nspp,
		int	nyrs,
		const vector< int > &	suit_styr,
		const vector< int > &	suit_endyr,
		const vector< int > &	nyrs_suit,
		int	msmMode,
		const vector< int > &	nsex,
		const vector< int > &	nages,
		const vector< int > &	suitMode,
		array< Type > &	diet_prop,
		array< Type > &	avgN_at_age,
		array< Type > &	weight_hat,
		array< Type > &	other_food_diet_prop )

MSVPA/empirical suitability, from stomach content and prey biomass.

For predators with suitMode = 0. stom_div_bio is the observed diet proportion divided by prey biomass (numbers in thousands x weight in kg = mt; the prey term is squared under msmMode = 2, Type III). That ratio is normalized over the modeled prey and "other food", averaged over suit_styr..suit_endyr, and held constant across all years.

A prey age whose average numbers approach zero with a positive diet proportion sends stom_div_bio, and so the prey term, to infinity: that one cell takes the whole normalized share, "other food" vanishes, available food collapses with it, so M2 diverges. The other-food term is floored at a fixed fraction of the prey term, bounding suit_other below at floor/(1 + floor), leaving sum(suitability) + suit_other = 1 exact. .warn_suitability_floor() reports when it binds.

Parameters

stom_div_bio	Ratio of diet proportion to biomass, updated; [nspp*max_sex, nspp*max_sex, max_age, max_age, nyrs].
suitability	Final calculated suitability, updated.
suit_other	Remaining suitability for "other food", updated.
nspp	Number of species.
nyrs	Years, hindcast plus projection.
suit_styr	Start year index of the suitability window, per predator.
suit_endyr	End year index of the suitability window, per predator.
nyrs_suit	Years in the suitability window, per predator.
msmMode	Predation mode; 2 (Type III) squares the prey numbers.
nsex	Sexes modelled per species: 1 = combined, 2 = both.
nages	Age bins per species.
suitMode	Suitability mode per predator species.
diet_prop	Observed diet proportions from organize_diet_obs().
avgN_at_age	Average numbers-at-age (thousands of fish).
weight_hat	Estimated weights-at-age (kg).
other_food_diet_prop	Residual diet proportion for "other food", per mt of it.

Definition at line 180 of file predation.hpp.

References imax(), and max2().

calculate_parametric_suitability()

template<class Type>

void calculate_parametric_suitability	(	array< Type > &	suitability,
		array< Type > &	suit_other,
		int	nspp,
		int	nyrs,
		const vector< int > &	nsex,
		const vector< int > &	nages,
		const vector< int > &	suitMode,
		array< Type > &	length_hat,
		array< Type > &	weight_hat,
		matrix< Type > &	vulnerability,
		const vector< Type > &	vulnerability_other,
		const vector< Type > &	gam_a,
		const vector< Type > &	gam_b )

Parametric suitability (Gamma, Lognormal, Normal) from size ratios.

Evaluates the predator-to-prey size ratio against a density scaled to a peak of 1. suitMode sets both the density and the size metric:

• Gamma: suitMode 1 (length-based) or 2 (weight-based)
• Lognormal: suitMode 3 (length-based) or 4 (weight-based)
• Normal: suitMode 5 (length-based) or 6 (weight-based) The length-based modes 1, 3 and 5 are NOT YET AVAILABLE: they need an active growth model, and data_check() blocks them until that validation is added.

Parameters

suitability	Size-based suitability, updated [nspp*max_sex, nspp*max_sex, max_age, max_age, nyrs].
suit_other	Baseline vulnerability to "other food", updated; [nspp, max_sex, max_age, nyrs] – species NOT collapsed, unlike suitability.
nspp	Number of species.
nyrs	Years, hindcast plus projection.
nsex	Sexes modelled per species: 1 = combined, 2 = both.
nages	Age bins per species.
suitMode	Parametric suitability mode per predator species.
length_hat	Lengths-at-age [nspp*2+n_flt, max_sex, max_age, nyrs].
weight_hat	Weights-at-age [nspp*2+n_flt, max_sex, max_age, nyrs].
vulnerability	Bounded predator-prey vulnerability [nspp, nspp].
vulnerability_other	Vulnerability preference for "other food" [nspp].
gam_a	Shape/mean parameter for predator size-selectivity [nspp].
gam_b	Scale/sd parameter for predator size-selectivity [nspp].

Definition at line 76 of file predation.hpp.

calculate_vulnerability()

template<class Type>

void calculate_vulnerability	(	matrix< Type > &	vulnerability,
		vector< Type > &	vulnerability_other,
		int	nspp,
		const vector< int > &	suitMode,
		matrix< Type > &	log_phi )

Transforms predator-prey preference parameters into vulnerability.

Multinomial logistic transform: each predator's prey preferences and its vulnerability to "other food" sum to 1. Parametric suitability only (suitMode > 0). Adopted from Trijoulet et al. 2020 (https://github.com/vtrijoulet/Multisp_model_JAE/blob/master/MS_SSM.cpp).

Parameters

vulnerability	Bounded predator-prey vulnerability [nspp, nspp], updated.
vulnerability_other	Remaining vulnerability for "other food" [nspp], updated.
nspp	Number of species.
suitMode	Suitability/preference mode per species.
log_phi	Estimated log-preference parameters [nspp, nspp].

Definition at line 20 of file predation.hpp.