A bioenergetics model for seasonal growth of Indian oil sardine (Sardinella longiceps) in the Indian west coast

HIGHLIGHTS

• Developed a sardine bioenergetics model coupled with a lower trophic level model (NEMURO).

• Model reproduced growth rate and wet-weight of sardine along the Indian west coast.

• Utilized model to describe the seasonality in the growth of sardine.

• Simulated wet weight validated with available observations.

• Sensitivity analysis revealed the importance of temperature and consumption in the growth.

Abstract

The Indian oil sardine (Sardinella longiceps) fishery was observed with wide stock fluctuation. Ecological parameters, mainly temperature and prey density, had a profound influence on the growth of S. longiceps and hence the production. In this study, a fish bioenergetics model, coupled with a lower trophic level model, was developed to reproduce the seasonality in the growth of S. longiceps. For this, we have used an Indian Ocean adaptation of an intermediate complex ecosystem model called North Pacific Ecological Modeling for Understanding Regional Oceanography (NEMURO) for marine productivity simulations. The model has 11-component ecosystem variables such as two types of phytoplankton (small and large including flagellates and diatoms), three types of zoo-planktons (small, large, and predatory, which includes ciliates, copepods, and euphausiids), particulate and dissolved organic matter, opal, cycling of nitrate, ammonia, and silicate. The prey densities derived from the NEMURO were input to the sardine bioenergetics model. The coupled model reproduced the appropriate growth rate and wet-weight of S. longiceps in its seasonal cycle in four major fishery regions such as Kochi, Goa, Ratnagiri, and Mumbai as verifiable from the available observation. In Kochi, the mean wet weight was 72.0 ± 12.8 g (June to September), 65.4 ± 5.3 g (October to November), 82.4 ± 2.7 g (December to February), and 66.7 ± 3.8 g (March to May). Goa and Ratnagiri have moderate weights with mean wet weight as 73.6 ± 10.6 g (June to September), 87.4 ± 3.2 g (October to November), 95.5 ± 4.3 g (December to February), and 76.2 ± 6.8 g (March to May). In the Mumbai region, maximum weight is simulated with mean wet weight as 97.4 ± 13.3 g (June to September), 102.1 ± 1.6 g (October to November), 104.8 ± 1.3 g (December to February), and 101.6 ± 1.2 g (March to May). Sensitivity analysis revealed the importance of temperature and consumption in the growth of sardine. Detailed model validation with available observations is presented.

Introduction

Indian oil sardine (Sardinella longiceps, after this referred to as SL) is the single most significant contributor to India's annual marine fish production. It is an important species with great value regarding the nation's economic and ecological aspects (CMFRI, 2018). Even though commercial exploitation of SL is mainly confined to the southwest coast of India, its distribution is found widespread along the entire coast of peninsular India, extending from the Gujarat coast in the Arabian Sea to the West Bengal coast in the Bay of Bengal (Longhurst and Wooster, 1990; Madhupratap et al., 2001; Kripa et al., 2018).

The fishery of SL commences during the early monsoon period (i.e., June-July) with the entry of spawners belonging to the various maturity stages to the inshore waters. This period is generally considered as the beginning of the SL biological year. During June-September, the peak spawning period is observed, which may occasionally be earlier or prolonged by another month (Balan, 1964; Antony Raja, 1969; 1970). After hatching, the larvae mouth is well-formed by the second day, and the yolk absorption is completed successively. It should coincide with the ‘critical first feeding’ for ensuring the survival of larvae (Hjort, 1914; Nair, 1973; Anon, 1976; Rohit et al., 2018). SL attains maturity by the end of the first biological year.

A rapid increase in length (i.e., about 70% of its first-year length) establishes during the first 2 to 3 months of the larval growth. Juveniles are recruited to the fishery from August to October (which is spawned from June to September). In a later period, the juveniles are dominated due to the entry of large shoals. By January and February, SL fishery is gradually reduced, and after that, fish may migrate from the coastal waters to the offshore. The next year, they re-enter the coastal waters for their second spawning along with the virgin spawners (Antony Raja, 1969, 1970).

During the last few decades, wide fluctuations in this species production along the southwest coast were observed (Krishnakumar et al., 2008; Kripa et al., 2018; Rohit et al., 2018; Hamza et al., 2021). Ecological parameters profoundly influence SL's growth and reproduction and hence stock fluctuations (Krishnakumar et al., 2008; Barange et al., 2009; Selvin and Lipton, 2012). Among these parameters, temperature and prey densities are the most important, determining the growth rate and metabolic activities of sardines (Matsuoka, 2001; Joseph and Jayaprakash, 2003; Morales-Bojórquez et al., 2003; Takahashi et al., 2009).

The growth of SL on the west coast of peninsular India is mainly studied using length-frequency-data and by analyzing growth inscriptions encoded in otoliths. However, life history observations of this particular species are limited as they are difficult to rear in laboratory settings. Under these circumstances, numerical models can act as a vital alternative to simulate pelagics growth, study the missing information, and complement the data-gap that arises due to limited direct observations.

To study the growth of SL and its seasonal cycle numerically, an appropriate numerical algorithm capable of capturing the relation and effect of various ecological parameters on the growth of fish species is required. Such models should further link the lower trophic level (LTL) to the critical fish species in the higher trophic level (HTL). Generally, LTL models are developed to analyze the biogeochemical cycling of nutrients and plankton productions. These models consider the HTL as only for mortality parameterization on the planktonic species without explicit numerical simulation (Fennel and Neumann, 2004). Conversely, most of the fish models begin with zooplankton food density as an external driving force, and less attention is given to the nutrient and LTL food-web connections (Rose et al., 1999; Vlymen, 1977; Beyer and Laurence, 1980; Letcher et al., 1996). In this perspective, a model that integrates the LTL-HTL linkage is a need-of-the-hour to study the growth and distribution of a fish species in a biologically meaningful way (Hinckley et al., 1996; Werner et al., 1996; Hermann et al., 2001).

The North Pacific Ecosystem Model for Understanding Regional Oceanography (NEMURO) is a well-known LTL model with an added fish bioenergetics module called NEMURO.FISH (Kishi et al., 2007; Megrey et al., 2007a). The LTL-NEMURO model has been applied to a variety of locations in the global oceans. It simulates the daily predator-prey interactions and biogeochemical cycling of nutrients (Yamanaka et al., 2004; Kishi et al., 2004). Its recent adaptation for the Arabian Sea is documented in Anju et al. (2020). Simultaneously the NEMURO.FISH is used to study the growth and population dynamics of a variety of commercially important species such as Pacific Saury (Cololabis saira), Pacific Herrings (Clupea harengus pallasi), and Japanese Sardine (Sardinops melanostictus) (Ito et al. 2004; Megrey et al. 2007b; Okunishi et al. 2009; Okunishi et al. 2012). This model is also used to simulate fish growth in different geographical locations and compare the environmental aspects and growth responses without considering migration (Rose et al., 2007; Megrey et al., 2007a).

A fish bioenergetics model connecting LTL-HTL linkage is not attempted to Indian Ocean regions for any fish species to the best of our knowledge. Therefore, as a first step, a fish bioenergetics model (i.e., NEMURO.FISH), coupled with an LTL-model (i.e., NEMURO), is used to investigate the relationships between sea surface temperature (SST), prey density, and SL growth at the southwest coast of peninsular India in the Arabian Sea. In our earlier study, the LTL NEMURO is implemented for the Arabian Sea coupled with a 1-D mixed layer model and simulated the predator-prey interactions realistically (Anju et al., 2020). In the present study, we extended this model to couple it with the fish bioenergetics model and reproduced SL's growth on the Arabian Sea's eastern coast.

The present study has the following objectives: (i) to develop a fish bioenergetics model for HTL species coupled with LTL species to suit the environmental conditions of the Arabian Sea, (ii) utilize the model to describe the growth of SL along the west coast of peninsular Indian different geographical locations (without any explicit migration), (iii) examine the seasonality in the growth of SL as simulated in the model and offer validation at selected fishing locations along the west coast of peninsular India, and (iv) investigate the sensitivity of the model results to various critical parameters including water temperature and prey density which are helpful in the fisheries management.