Feeding kernels from stomach distributions
Gustav Delius
05/11/2021
Source:vignettes/feeding_kernels.Rmd
feeding_kernels.RmdFeeding kernel
Now we will discuss how the stomach content is related to the feeding kernel.
The stomach content is the result of a balance between the rate at which prey items are ingested and the rate at which they are digested. So we start by looking at ingestion and digestion rates.
Ingestion
The rate density at which a predator of size swallows prey of size is proportional to the product of the feeding kernel and the prey number density : This is a rate density, fin the sense that if one integrates this over a range of prey sizes one gets the rate at which the predator swallows individuals from this size range.
The proportionality constant only depends on the predator size and, as we will see, we will not need it in our analysis. The prey number density is defined as usual such that is the total number of potential prey items with sizes between and .
In a multi-species model in which the predator has different interaction strengths with different prey species, given by an interaction matrix , the prey number density for a predator of species would be obtained from the number densities of individual prey species as Of course, we do not know the prey density that the predators experienced that were sampled in the stomach observation. So the best we can do is to assume that it was reasonably close to the Sheldon spectrum with .
Digestion
Prey items will stay identifiable in the predator stomach only for a certain period of time before they disintegrate. After that time they will not contribute to the observation of prey items in the stomach. Of course we do not know in detail the rate at which prey items are digested. However it is reasonable to make the approximation that the rate at which a prey item has its body mass digested away scales with body size as because digestion acts on the surface of the prey item and this surface scales approximately as . Let us assume that a prey item becomes unidentifiable when a fixed percentage of its body mass is digested. The time until that has happened is proportional to the body mass divided by the rate at which mass is digested: . The rate at which a prey item of size disintegrate is equal to the inverse of the disintegration time. Denoting that rate by we have The proportionality constant will not be important in what follows.
Balance equation
We describe the size distribution of prey in stomachs of predators of size by the density function . See the Density functions vignette for a discussion of this density function.
The observed stomach content is such that the rate at which prey items of a particular size are ingested is equal to the rate at which such items disintegrates due to digestion and thus . Using our expressions for these rates we obtain This we can solve for the feeding kernel: Using further that (see Density functions) where and , we find This tells us that if the stomach distribution is described by the truncated exponential distribution with rate , the feeding kernel as a function of is described by the truncated exponential distribution with rate .
If the stomach distribution is the normal distribution with mean and variance then the feeding kernel as a function of is also a Gaussian with the same variance and with mean .
We assume that on average the stomach observations reflect the steady state with . We now substitute our expressions for the various rates into the resulting balance equation: