Allow for non-feeding larval stage

[gustavdelius]

For most marine species there is a non-feeding larval stage that can last a few weeks. Thus individuals stay in their smallest size class longer than currently modelled. To take this into account, we should decrease the growth rate in the smallest size class so that the time it takes to grow into the next size class is increased by the length of this non-feeding stage. This will lead to lower survival due to longer exposure to high mortality. This mortality is density-dependent, thus increasing the non-linearity in the emergent stock-recruitment relationship.

Currently the time it takes an individual to grow into the second size class is dw / g where dw is the width of the first size class and g is the growth rate in the first size class. We want to increase that time by T where T is the time the larvae need before they start feeding. We can achieve this by decreasing the growth rate by a factor of 1 + T g / dw.

Strictly speaking when we decrease the growth rate to mimick the non-feeding stage, we should also decrease the rate at which the food for the larvae is used up. But this effect on the food resource will be negligible because the consumption from a single size class is small.

So this should be easy to implement in code. We would write a function nflsEGrowth() that first calls the existing mizerEGrowth() and then multiplies the result in the smallest size class for each species by the factor of 1 + T g / dw before returning that modified growth rate, where the value for T would come from a new column in the species_params data frame (any suggestion for a good name for this parameter?). The user who wants to use this would have to register this new function with params <- setRateFunction(params, "EGrowth", "nflsEGrowth").