Model of Zooplankton Optimal Current Feeding9

MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog SerVol. 403: 129–144, 2010

doi: 10.3354/meps08466Published March 22

INTRODUCTIONMarine zooplankton forms the first gateway in theprocessing of organic matter derived from primaryproduction in the surface ocean. The fate of thisorganic matter is thus to a large part a function of 3quantities which define zooplankton feeding behav-iour and growth: (1) specific ingestion rateIis the totalfood intake, (2) assimilation efficiency Edefines theassimilated and excreted fractions of the ingestedmaterial, and (3) respiration and exudation separateremineralisation from growth, i.e. accumulation of bio-mass available for utilisation by higher trophic levels.Net growth rate gcan be defined as the balancebetween assimilation EIand respiration R:g= EI– R(1)where Rrepresents respiration of CO2and exudationof nutrients (all symbols are summarised in Table 1).Sloppy feeding and risk of higher predation will not beconsidered in the present study. Excretion in the formof faecal pellets can be exported to the deep ocean andthereby contribute to the biological carbon pump. Con-sequently, the zooplankton formulation strongly af-fects the behaviour of plankton models both directlyvia its role in the food web and indirectly via its impacton vertical nutrient profiles (Steele & Henderson 1992).Nevertheless, although many zooplankton modelsexist (Gentleman et al. 2003), they focus almost exclu-sively on ingestion and do not usually allow fordynamic adjustments in feeding behaviour (Paffen-höfer et al. 2007).Behaviours effecting prey encounter can broadly becategorised into several foraging strategies. The sim-

© Inter-Research 2010 · www.int-res.com*Email: mpahlow@ifm-geomar.de

Model of optimal current feeding in zooplanktonMarkus Pahlow*, A. E. Friederike ProweIFM-GEOMAR, Düsternbrooker Weg 20, 24105 Kiel, Germany

ABSTRACT: Zooplankton feeding formulations in plankton models have exclusively focused on therelation between food concentration and ingestion, with respiration and excretion being treated sep-arately, despite experimental evidence for strong links among these processes. We present an opti-mal current-feeding model linking ingestion, respiration, and assimilation efficiency to foragingactivity. The Ivlev model is a special case of our optimal current-feeding model, which applies tostatic feeding behaviour. We validate our model with experimental data for copepods, ciliates, anddinoflagellates. Parameter estimates suggest that phylogenetic grouping is more important than pre-dator size in determining feeding behaviour. Respiratory costs of foraging, e.g. for generating a feed-ing current, may be much larger than previously thought, are larger in smaller organisms, and mightexplain the independent development of feeding thresholds in different micro- and mesozooplank-ton groups. Both preferential feeding on, and lower feeding thresholds for, larger food particles arepredicted to derive from greater capture efficiency owing to enhanced detectability of larger parti-cles. The relation between feeding threshold and prey size appears to depend on feeding strategy butnot on predator size, as a common relationship seems to apply for current feeders (ciliates and cope-pods) spanning a vast size range. Our model exhibits an inverse relationship between ingestion andassimilation efficiency, reducing the contribution of copepods to exportof organic matter relative toremineralisation at low food concentrations. Export ratio variations previously thought to requirestrong shifts in community composition can be generated by changes in feeding behaviour predictedby our model.

KEY WORDS: Zooplankton · Optimal foraging · Current-feeding modelResale or republication not permitted without written consent of the publisherMar Ecol Prog Ser 403: 129–144, 2010plest possible behaviour is simply waiting for randomprey encounters (ambush feeding), which is efficientfor moving prey only (Gerritsen & Strickler 1977). Themost effective means of increasing prey encounters iscruise feeding, i.e. swimming in various patterns, butin addition to metabolic costs, this strategy also in-evitably increases encounters with higher predators(Visser et al. 2009). A feeding current can enhanceprey encounters as well, albeit somewhat less effec-tively, owing to its limited spatial extent. However,because the predator itself is not moving, this strategydoes not suffer the increased risk of higher predationincurred by cruise feeding. Filter feeders direct theirfeeding currents through a filter, sieving out prey par-ticles, which works most efficiently with very large fil-ters (Gerritsen et al. 1988). Suspension-feeding cope-pods have often been considered filter feeders (e.g.Conover 1968, Frost 1972, Lehman 1976, Vidal 1980a),but one important yet still often ignored aspect of thisfeeding mode is that a copepod’s feeding current doesnot pass through but actually around its feeding ap-pendages (Koehl & Strickler 1981), which actto detect and capture prey organisms from thefeeding current (Visser & Stips 2002). There-fore, we refer to this feeding mode as currentfeeding.Grazing functions commonly applied inplankton models include the Holling types,based on the disk equation, which in turn wasderived for insect feeding (Holling 1959, 1973),and the Ivlev equation, derived for fish (Ivlev1961). Their application to zooplankton wasempirically motivated (Conover 1968, Fujii etal. 1986) and had no mechanistic basis. Asidefrom the inappropriateness of the term filterfeeder to describe suspension feeding by cope-pods, this lack of a mechanistic foundationmeans that it is difficult to generalise thesemodels and analyse disagreements betweenmodels and observations. For example, Parsonset al. (1967) noted the necessity to introduce afeeding threshold into the Ivlev equation in or-der to describe the behaviour of copepods. Vi-dal (1980a) and Kiørboe et al. (1982, 1985) in-troduced entirely new empirical formulationsto describe their observations, which fit theirdata better but could ‘not be interpreted in bio-logical terms’ (Kiørboe et al. 1982, p. 185).Almost all existing zooplankton feedingmodels describe ingestion as a function of foodconcentration, without any inherent link torespiration or assimilation efficiency, despitemany observations that these processes covaryin a systematic manner (e.g. Landry et al. 1984,Kiørboe et al. 1985). Respiration can be consid-ered the sum of 3 components: (1) a constant mainte-nance respiration thought to sustain standard (resting)metabolism, (2) the cost of foraging, and (3) the meta-bolic cost of assimilation, termed specific dynamicaction (Steele & Mullin 1977). The increase of respira-tion with increasing growth rate is usually attributedmostly to specific dynamic action (Steele & Mullin1977, Kiørboe et al. 1985), whereas the cost of foraginghas been suggested to give rise to feeding thresholdsin order to save energy in the absence of food (Frost1975). Zooplankton assimilation efficiency appears todecrease with increasing growth rate (Steele & Mullin1977, Kiørboe et al. 1985), yet it is usually treated asconstant in plankton models (Lam & Frost 1976, Steele1998). The only existing feeding model utilising mech-anistic relationships among ingestion, respiration, andassimilation efficiency is the optimal foraging descrip-tion of filter feeding by Lehman (1976). However, thismodel has never been validated and, because it cannotbe written in closed form, is not suitable for use inlarger plankton models.130SymbolStandard unitsDescriptionAfd–1Specificforaging activityAf,maxd–1Maximum specific foraging activityAtd–1Maximum total specific activityα–Prey handling coefficientβ–Digestion (assimilation) coefficientca–Cost of assimilation coefficientcf, cf*–Cost of foraging activity coefficientaE–Assimilation efficiency~Eg–Modified gross growth efficiency