Hypoxia Effects Within an Intra-guild Predation Food Web of Mnemiopsis leidyi Ctenophores, Larval Fish, and Copepods

Abstract

Differences in predator and prey tolerances to abiotic factors, such as seasonal low dissolved oxygen (DO) concentrations in estuarine environments, can affect planktonic food web dynamics. Summertime hypoxia in the Chesapeake Bay alters field distributions, encounter rates, and predator–prey interactions between hypoxia-tolerant ctenophores, Mnemiopsis leidyi, and their less tolerant ichthyoplankton and zooplankton prey. Omnivory and intra-guild predation (IGP) increase the complexity of food webs, thereby confounding the effects of predation versus competition on prey populations. Omnivorous ctenophores in temperate estuarine food webs can both eat and compete with fish larvae for copepod prey. We isolated the effects of predation and competition, and how low versus high DO, affect larval fish growth and survival, using a spatially explicit (three vertical layers) individual-based model of a ctenophore-fish larvae-copepod IGP food web. We simulated three alternative food web structures of how ctenophores affect fish larvae (full interactions, relaxed predation, relaxed competition) under normoxic and hypoxic DO scenarios. Results from laboratory experiments and field studies were used to configure and corroborate the model. Ctenophore predation had a bigger effect on survival of modeled fish larvae than did competition between ctenophores and fish larvae for shared zooplankton prey, but competition more strongly affected larval fish growth rates than did predation. Hypoxia versus normoxia did not alter the relative importance of ctenophore predation and competition, but low DO did decrease larval fish survival and increase larval growth rates. Model results suggest that consideration of the interaction strength in food webs and explicit treatment of spatial habitats to allow predator–prey overlap to emerge from movement will enhance our ability to predict hypoxia effects on fish.

Keywords

Appendix A. Stage-Based Matrix Projection Models for Fish Eggs and Yolk Sac Larvae, and Ctenophore Eggs and Larvae

Six stage-based matrix projection models were used to update fish eggs and yolk sac larvae, and ctenophore eggs and larvae. There was a model for fish and for ctenophores for each of the three layers. The models were 2 × 2 and operated on a 12-h time step. For each model, we computed the diagonal and subdiagonal elements from stage survival and duration every 12 h. We first computed from survival over 12 h for the ith stage from specified daily instantaneous survival rates as S _i  = e ^{−M /2} _i . For fish eggs only, DO was used to compute SurEggDO (Eq. 11.13), and S _i for eggs was then adjusted as S _i ·SurEggDO. We then computed φ_i, survival for each time period, from S_i and duration (D_i, number of 12 h):

$$ \varphi_{i} = \frac{{S_{i}^{D} - S_{i}^{D - 1} }}{{S_{i}^{D} - 1}} $$

The diagonal and off-diagonal elements were then:

$$ D_{i,i} = S_{i} \cdot \left( {1 - \varphi_{i} } \right) $$

$$ D_{i,i + 1} = S_{i} \cdot \varphi_{i} $$

Fecundity (usually the top row of the matrices) was dealt with by simply adding newly entering eggs to those already present in each layer every 12 h. Number of eggs added was computed based on day of year, and dynamically each 12 h for ctenophores based on growth and summed over individual ctenophores.

At the beginning of each 12-h time step, the matrices were specified and the numbers of individuals in each stage were updated. Newly entering eggs for fish and ctenophores were then added to their egg abundances. Then during the next 12 h, consumption of fish eggs and yolk sac larvae by ctenophores was subtracted from the total number of individuals in each layer. The decreased numbers of individuals in each life stage in each layer were then used to start the next time step.

The mortality rates, durations, and fecundity rates are shown in Table 11.7. Typical matrices for each of the taxa were:

Table 11.7 Mortality, stage duration, and fecundity rates for stage-based matrix projection models for fish eggs and yolk sac larvae, and ctenophore eggs and larvae

Fish:

$$ \left[ {\begin{array}{*{20}c} {0.75} & 5 \\ {0.25} & {0.75} \\ \end{array} } \right] $$

Ctenophores:

$$ \left[ {\begin{array}{*{20}c} {0.402} & {20} \\ {0.27} & {0.402} \\ \end{array} } \right] $$