Ontogenetic changes in the food items assimilated by Plagioscion squamosissimus (Perciformes: Sciaenidae) and Hypophthalmus edentatus (Siluriformes: Pimelodidae)

Abstract

In order to identify the ontogenetic changes in the food items assimilated by the larvae of Plagioscion squamosissimus and Hypophthalmus edentatus, we analyzed the stomach content and the nitrogen isotopic value (δ ¹⁵N) from the larvae. Samplings were performed monthly, from October 2002 to March 2003, and October 2003 to March 2004 in the subbasin of the Ivinheima River (Upper Paraná River floodplain), Mato Grosso do Sul State, Brazil. The food items assimilated by the larvae of P. squamosissimus changed with respect to the developmental stages. In the preflexion stage, we found a large contribution of phytoplankton, followed by zooplankton. During the flexion and post-flexion stages, we recorded a decrease in phytoplankton contribution, an increase in zooplankton contribution, and the presence of zoobenthos. In H. edentatus, phytoplankton made the greatest contribution, followed by zooplankton. Ontogenetic changes in δ ¹⁵N for this species were not detected. For both species, the food item predominantly ingested was not the assimilated one, except for the last two developmental stages of P. squamosissimus.

Keywords:

• larvae
• stomach
• fish
• δ¹⁵N

Introduction

Understanding the events that ensure the survival of ichthyoplankton depends on the comprehension of ontogenetic shifts in the diet, which is helpful for monitoring fish stocks. Ontogeny is a continuous process that includes transitory accelerations rather than just a predictable succession of the ontogenetic stages (Kamler 2002). In the larval period there are intense morphological transformations leading to changes in diet or even in habitat utilization. The larval size influences these changes, the capacity to use resources, and the character of the interaction with other individuals (Werner and Gelliam 1984).

A predator's food items can be quantified by the analysis of the stomach contents in terms of the specific taxon ingested; the same is not necessarily true of material that is assimilated (Grey et al. 2002). Nitrogen isotopic values can be used to make inferences about the diet diversity and the trophic position of organisms (Gorokhova et al. 2005). Since their values increase with each trophic level transfer, this analysis has the advantage of revealing the item assimilated among those ingested. The combination of stomach content and stable isotope analyses has become a complementary technique (Grey et al. 2002) and was used by Fisher et al. (2001), Grey et al. (2002), and Jones and Waldron (2003) to evaluate food items ingested and assimilated during the ontogenetic changes of fish.

The most frequently encountered fish larvae in the upper Paraná River floodplain are those of introduced Plagioscion squamosissimus and the native Hypophthalmus edentatus (Agostinho et al. 2004). In adult form, the P. squamosissimus was considered piscivorous by Hahn et al. (1997a) and contributes to the reduction of abundance of H. edentatus through predation (Ambrósio et al. 2001). Lansac-Tôha et al. (1991) and Hahn et al. (1997b) classified H. edentatus adults as a planktivorous, using zooplankton as their predominant food source.

For most species, the larval ecological requirements are distinct from those of the adults from the same species (Nakatani et al. 2001). Therefore, in order to test this assumption, we investigated ontogenetic changes in the food items assimilated by larvae of P. squamosissimus and H. edentatus through stomach content analysis and using the nitrogen isotopic values from these items. We also examined whether the predominantly ingested item was also the predominantly assimilated one for each species.

Materials and methods

Samples were collected monthly at the Finado Raimundo lagoon and the Ivinheima River, from October 2002 to March 2003 and October 2003 to March 2004, during the spawning season of the fish species. We used a conical–cylindrical plankton net to sample during the nychthemeral cycles, with 4-hour intervals between sample collections, over a period of 48 h.

The most abundant species in the samples were P. squamosissimus and H. edentatus. After fixation in 70% ethanol, the specimens were separated into the developmental stages of preflexion, flexion, or postflexion stages, according to their notochord flexion and the terminology described by Ahlstrom et al. (1976) and modified by Nakatani et al. (2001). Among the 3100 individuals, 196 larvae were randomly selected for analysis of stomach content. These larvae were grouped into standard length (SL) classes, according to their developmental stages. The P. squamosissimus samples, comprising 115 individuals, were divided into the following four classes, based on their size and their developmental stages: preflexion (3.1–5.0 mm); preflexion and flexion (5.1–7.0 mm); flexion (7.1–9.0 mm); and postflexion (9.1–11.0 mm). The H. edentatus samples, comprising 81 individuals, were distributed into the following three size classes: preflexion (3.5–7.5 mm), flexion (7.51–11.5 mm), and postflexion (11.51–15.5 mm). Some P. squamosissimus larvae at the preflexion and flexion stages had to be grouped together due to the SL range observed in some individuals.

In the preflexion and flexion larvae, the entire content of the digestive tube was analyzed, whereas in the postflexion larvae, only two-thirds of the digestive tube contents were analyzed because of the advanced degree of digestion observed in food items in the hind portion of the tract.

In order to determine nitrogen isotopic values, we had to group 8–10 individuals to obtain a 0.5 mg sample for each development stage. The digestive tube from each specimen was removed to avoid interference in the isotopic composition of the fish tissues analyzed. The specimens were rinsed with distilled water to remove excess alcohol and dried at 60°C. The same samples used here had their isotopic results corrected by subtracting 0.5‰. This correction value for ¹⁵N was obtained by Manetta et al. (in press) for the same species in the adult phase and recorded by Vander-Zanden et al. (2003) and Pease et al. (2006) for adults and larval marine fishes, respectively. Afterward, the samples macerated until a fine powder was obtained. They were then sent to Center for Nuclear Energy in Agriculture – USP/SP for nitrogen isotopic ratio determination.

To calculate contributions of food item in the diets of P. squamosissimus and H. edentatus, the Moore–Penrose pseudo-inverse matrix method was used. The method consisted of the determination of the various prey types (f) by inverting a 2 × 2 inverse matrix, whose elements are one (1) and the δ ¹⁵N_mix value for different prey types (f) (Hall-Aspland et al. 2005). In P. squamosissimus, the main groups identified in the larval diet were phytoplankton (f _a), zooplankton (f _b), and zoobenthos (f _c), except for the preflexion stage, when the larvae only ingested f _a and f _b. In H. edentatus, the largest groups identified in the larval diet were f _a and f _b. A 3.4‰ value was added to the mean nitrogen isotopic value of each food item, corresponding to the assumed trophic level for the predatory species when ingesting a given prey item (Post 2002). In this case, the corrected values of prey used in the predator calculations were f _a = 6.7‰, f _b = 10.08‰ and f _c = 6.6‰. These isotopic values of the prey were collected at the time of this study because this work was part of a long-term ecological research project with several studies occurring at the same time. In order to identify the actual contribution of the item in the diet throughout development, the vector equation specified below was established to replace the mean larval nitrogen value (δ ¹⁵N_mix) for each development stage.

The equation that relates the δ ¹⁵N_mix predator isotopic value to the isotopic value of its food items is given by ¹⁵N_a +  ⁵N_b + ··· +  ¹⁵N_n = δ ¹⁵N_mix. Where δ ¹⁵N_a and δ ¹⁵N_b are the isotopic signs considered for the prey items (Hall-Aspland et al. 2005).

Results and discussion

In the beginning of exogenous feeding, the larvae of P. squamosissimus and H. edentatus showed high percentage and numeric compositions of zooplankton in the digestive tract, which remained all during the ontogenetic development (Table 1). The greatest contribution of assimilated items for P. squamosissimus, during the preflexion came from the phytoplankton, while during flexion and postflexion stages it came from zooplankton. As verified that as the larvae developed, the zooplankton and zoobenthos contribution increased, while the phytoplankton contribution decreased. For H. edentatus, the phytoplankton was the assimilated item of larger contribution in all the developmental stages. For both species, the food item predominantly ingested was not the principal item assimilated, except for the last two developmental stages of P. squamosissimus (flexion and postflexion).

Table 1. The percentage composition of ingested food based on numeric abundance and δ¹⁵N signature and the percentage of that food assimilated in tissue based on δ¹⁵N signature by larval stages of P. squamosissimus and H. edentatus.

	Preflexion			Flexion			Postflexion
	Numeric abundance	Ingested food item	Larval tissue	Numeric abundance	Ingested food item	Larval tissue	Numeric abundance	Ingested food item	Larval tissue
P. squamosissimus
Phytoplankton	46	27	53	48	22	22	25	9	20
Zooplankton	54	71	47	51	74	56	74	84	61
Zoobenthos	0	0	0	0	3	21	1	7	19
H. edentatus
Phytoplankton	44	20	57	42	16	57	44	14	65
Zooplankton	56	79	42	58	82	43	56	87	35

For P. squamosissimus, the assimilated food sources (δ ¹⁵N) were the phytoplankton, zooplankton, and zoobenthos. Plagioscion squamosissimus showed a gradual increase in δ ¹⁵N with size, suggesting an ontogenetic diet change. Food sources enriched in δ ¹⁵N gradually increased the importance of the zooplankton in the diet of the larvae.

For H. edentatus larvae, we verified that the assimilated food sources were the phytoplankton and zooplankton with the greatest assimilation of algae during larval development. Although this species did not show ontogenetic alteration, this result could be related to the morphological rapid development of the gill rakers that are longer and more numerous on the first arch and decrease toward the fourth arch. This is related to the planktivorous habit and filter feeding that is defined during the juvenile period (Makrakis et al. 2005).

The highest percentage of ingested items (zooplankton) was different from the assimilated one (phytoplankton) for the preflexion stage of P. squamosissimus and during the larval development of H. edentatus, and it may be related to the fast digestion of the algae in the digestive tube, making impossible the identification and the quantification of stomach content. Jomori et al. (2008) affirmed that the main source of food for marine and freshwater fish larvae is the zooplankton. Thus, the effects of predation by larvae on the zooplankton may result in alterations in the diversity and the density of species of zooplankton, as well as in phytoplankton composition and biomass and in the physical and chemical conditions of the environment (Soares and Hayashi 2005). For both species analyzed, the first stage of larval development exhibited a larger assimilation of the phytoplankton, which was persistent for all the other stages of the species planktivorous.

The feeding behaviors of the piscivorous and the planktivorous fish larvae are different. There are differences regarding their attack strategy against prey and their feeding capacity. The larvae of piscivorous fish are larger, exploit a greater volume of water per unit time, and pursue their prey until it is captured (Gerking 1994).

Post and Kitchell (1997) reported that the zooplanktivorous fish larvae predominantly have a limited trophic ontogeny, and may spend their entire lives with this feeding habit. Nevertheless, piscivorous fish larvae exhibit a rapid ontogeny due to the presence of alternative prey and competition for larger zooplankton components.

The association of the techniques of isotopic analysis and stomach content analysis, adopted in this study, demonstrated the existence of changes in nutritional requirements in the ontogenetic development of fish larvae. Additionally, the assimilation of zooplankton and phytoplankton may not reflect, in general, the ingested proportions.

Acknowledgements

We thank PEA, NUPELIA, and PELD for providing financial and logistic support; CNPq, for making this project possible; Maria Salete Ribelato Arita, João Fabio Hildebrandt, and Márcia R. Paiva for providing assistance with the bibliographic material; Dr Liliana Rodrigues, Fábio Amadeo Lansac-Tôha, and Daniela Peretti for giving manuscript suggestions; and the entire team at UEM/NUPELIA's Laboratory of Ecology Energetic, Ichthyoplankton, and Zooplankton.