Oceanographic sampling methods and research cruises

In short

Oceanographic sampling is the set of instruments and procedures used to collect water, sediment, biota and physical measurements from the sea in a planned, repeatable way, and a research cruise is the voyage on which that plan is executed station by station. The quality of every downstream result - a salinity profile, a chlorophyll value, a benthic diversity index - is fixed at the moment of collection, so sampling design and gear handling are core examinable content. In India this knowledge sits behind the Ministry of Earth Sciences (MoES) fleet operated through the National Institute of Ocean Technology (NIOT) and the Centre for Marine Living Resources and Ecology (CMLRE), the CSIR-National Institute of Oceanography (NIO), INCOIS, the Fishery Survey of India and the Indian Navy's hydrographic service. It is examined in B.Sc/M.Sc Marine Biology and Oceanography practicals and vivas, in CSIR-UGC NET (Earth, Atmospheric, Ocean and Planetary Sciences), IIT JAM, GATE XL/GG where ocean sampling appears, and in technical rounds for research assistant, observer, survey and EIA-consultant posts.

Explained

Sampling design: the plan before the gear

• Station: a fixed geographic point (latitude, longitude) where a defined set of measurements is made; a transect is a line of stations, usually perpendicular to the coast to capture the land-sea gradient.
• Grid design: stations placed on a regular lattice (good for mapping and interpolation); stratified random, where the area is divided into strata (depth zones, habitat types) and stations randomised within each - the standard for estimating a mean with a defensible error.
• Nested/adaptive design: extra stations added near a feature found during the cruise (a front, an upwelling plume, an oxygen minimum).
• Standard depths for water sampling (a long-standing international convention): surface, 10, 20, 30, 50, 75, 100, 150, 200, 250, 300, 400, 500 m, then every 500 m. Always sample the thermocline and the chlorophyll maximum, whatever the standard list says.
• Replication: at least three replicate grabs or net hauls per station for any biological statistic; one grab gives no variance and no confidence interval.
• Time scales: a synoptic survey treats the ocean as frozen in time; a time-series station (repeat visits) separates seasonal signal from spatial pattern; a diel study (24 h, samples every 2-4 h) is needed for vertical migration of zooplankton.
• Pilot survey first: estimate variance, then compute the number of samples needed for a target precision.

Physical sampling and profiling

• Reversing thermometer with Nansen bottle (historical standard): the bottle is lowered open on a wire, a brass messenger trips it, it turns over, closes and reverses the mercury thermometer, freezing the reading. Accuracy about ±0.02 °C; now a teaching and backup method.
• CTD (Conductivity-Temperature-Depth): the workhorse. Conductivity cell plus platinum resistance thermometer plus strain-gauge pressure sensor, logging continuously as it is lowered at about 0.5-1 m/s. Salinity is computed from conductivity, temperature and pressure (practical salinity scale), never measured directly.
• Rosette (carousel) sampler: a frame carrying 12-24 Niskin bottles around a CTD; bottles are closed electronically on command at chosen depths during the upcast (the downcast gives the undisturbed profile). Add-on sensors: dissolved oxygen (Clark/optode), fluorometer (chlorophyll), transmissometer/turbidity, PAR, pH, nitrate (UV), altimeter.
• Niskin bottle: open-ended PVC tube with elastic-closed end caps; an internal spring or elastic snaps both caps shut. Teflon-coated or GO-FLO bottles are used for trace metals to avoid contamination.
• XBT (expendable bathythermograph): a cheap thermistor probe launched from a moving ship, data up the thin copper wire; gives temperature versus depth (depth inferred from a fall-rate equation, not measured) to about 760 or 1,800 m. No salinity, no recovery.
• ADCP (Acoustic Doppler Current Profiler): measures current velocity profiles from the Doppler shift of sound backscattered by particles; hull-mounted (underway), lowered (LADCP) or moored.
• Current meters and moorings: a subsurface mooring line with anchor, current meters, thermistors and an acoustic release recovered months later; the standard for long time series.
• Secchi disc: 20-30 cm white or black-and-white disc lowered until it disappears; the Secchi depth is a simple transparency measure and is converted approximately to the euphotic depth.
• Argo floats: autonomous profiling floats that park near 1,000 m, profile to about 2,000 m roughly every 10 days and transmit by satellite; India contributes and INCOIS hosts the Indian Argo data centre.
• Satellite remote sensing (sea-surface temperature, ocean colour, altimetry) gives synoptic coverage of the surface only, so it is validated against ship data - the ship provides "sea truth".

Chemical and biological water sampling

• Dissolved oxygen: drawn first from the Niskin bottle, with the tube to the bottom of the bottle, overflowing two to three volumes, no bubbles; fixed on deck with manganous chloride and alkaline iodide for the Winkler titration.
• Order of drawing from one bottle: dissolved gases (O₂, then any gas samples), pH/alkalinity, nutrients, chlorophyll/biological, salinity last - because the gas samples are the ones a delay or a bubble destroys.
• Nutrients: filtered (0.45 µm membrane, or 0.7 µm GF/F) into clean bottles, frozen at −20 °C or poisoned with mercuric chloride; analysed for nitrate, nitrite, phosphate and silicate by spectrophotometric (segmented-flow) methods.
• Chlorophyll-a: a known volume filtered on GF/F filters, filter frozen dark, pigments extracted in 90% acetone and measured fluorometrically or spectrophotometrically.
• Plankton nets: mesh size defines what is caught - about 20 µm for phytoplankton, 100-200 µm for mesozooplankton, 330-500 µm for larger zooplankton and fish larvae. Fitted with a flowmeter at the mouth to get the volume filtered.
• Net types: simple ring (Heron-Tranter, WP-2) for vertical hauls; Bongo (paired nets, two mesh sizes); Multiple Opening-Closing Net (MOCNESS / Multinet) for discrete depth strata; Bioness/Isaacs-Kidd midwater trawl for micronekton; Continuous Plankton Recorder towed from ships of opportunity.
• Haul types: vertical (depth-integrated, quantitative), horizontal (one layer), oblique (through the water column while steaming - the standard for fish eggs and larvae).
• Preservation: zooplankton in 4-5% buffered formaldehyde (hexamine or borax buffer, else carbonate structures dissolve); phytoplankton in Lugol's iodine; genetic samples in ethanol or frozen, never in formalin.
• Benthic fauna: sieved on a 0.5 mm sieve for macrofauna (0.3 mm in some protocols), 63 µm for meiofauna, stained with rose bengal, preserved in formalin.

Sediment and seabed sampling

• Grabs (surface, 0-10 cm, disturbed): Van Veen (long arms, good on sand and mud, about 0.1 m² or 0.2 m²), Petersen (heavy, coarse sediment), Smith-McIntyre (spring-loaded, works in rough weather), Ekman/Birge (soft mud, shallow water).
• Corers (retain stratification): gravity corer (free-fall, 1-3 m), piston corer (up to tens of metres), box corer (large undisturbed block, best for benthic ecology), multicorer (several small parallel cores, keeps the sediment-water interface intact - the standard for geochemistry), vibrocorer for sands.
• Dredges and trawls: naturalist's/rock dredge, beam trawl and Agassiz trawl for epifauna - qualitative, because the area swept is poorly known.
• Acoustics: single-beam echosounder (depth under the keel), multibeam (a swath, giving full-coverage bathymetry), side-scan sonar (acoustic image of seabed texture), sub-bottom profiler/chirp (shallow sediment layers), seismic survey (deep structure).
• Vehicles: ROV (remotely operated, tethered, with manipulators - sampling and video), AUV (untethered, pre-programmed survey), towed camera/TV sledge, and crewed submersibles; India's Samudrayaan/Matsya 6000 programme is building a crewed deep-sea submersible under the Deep Ocean Mission.
• Positioning: DGPS fixes the ship; the gear itself is positioned acoustically (USBL/transponder) in deep water, because wire-out is not depth when there is drift.
• Station log: every deployment is logged - station number, date, time (UTC), latitude, longitude, water depth, gear, wire-out, start and end time, sea state, operator. A sample without its log entry is scientifically worthless.

The research cruise: planning to data

1. Proposal and ship time: scientific objectives, station list, gear, personnel and berths; in India, ship time on the MoES fleet is allotted through NIOT/CMLRE and on NIO vessels through CSIR, usually a year ahead.
2. Permissions: survey inside another state's EEZ needs marine scientific research consent under UNCLOS Part XIII (India ratified UNCLOS in 1995); work in Indian waters may need clearances from the Ministry of Home Affairs and Ministry of Defence, Coast Guard intimation, and for protected areas permits from the Chief Wildlife Warden under the Wild Life (Protection) Act, 1972; CRZ and the Coastal Regulation Zone Notification, 2019 apply to coastal work; biodiversity access and benefit sharing under the Biological Diversity Act, 2002 (as amended 2023) applies to collecting biological resources; sampling in coral and marine protected areas is restricted.
3. Mobilisation: gear loaded, lashed for sea, lab containers plumbed, chemicals (formalin, acetone, acids) stored and declared, spares and consumables doubled - there is no chandler at sea.
4. Cruise plan: station sequence optimised for steaming time, weather window and daylight constraints (some operations need daylight); an alternative plan for bad weather. Steaming speed about 10-12 knots for a typical research vessel.
5. At sea: 12-hour watches, pre-deployment toolbox talk, hard hats, lifejackets and no hands on a loaded wire; CTD and net hauls on winches with load monitoring; samples logged, labelled (station, depth, date, gear) inside and outside the container.
6. Underway data: thermosalinograph, hull ADCP, echosounder, meteorological mast and fluorometer run continuously along the track.
7. Demobilisation and data: samples offloaded under cold chain, instruments calibrated post-cruise, a cruise report written (track chart, station table, preliminary results) and data submitted to the national archive - the Indian National Centre for Ocean Information Services (INCOIS) hosts the Indian Ocean data centre and NIOT/CMLRE hold cruise archives.
8. Safety and statutory frame: Indian-flag research vessels are registered and surveyed under the Merchant Shipping Act, 2025 (which replaced the 1958 Act from 15 March 2026; the regulator, formerly the Directorate General of Shipping, is now the Directorate General of Maritime Administration), crew certification under STCW, and SOLAS/MARPOL for safety and discharge. Shore-side and shipyard occupational safety now falls under the Occupational Safety, Health and Working Conditions Code, 2020. Discharge of chemical waste at sea is governed by MARPOL Annex V/II as given effect in Indian law - formalin and acetone waste comes back to port.

India's research vessels and programmes

• ORV Sagar Kanya (1983, about 100 m, MoES) - multidisciplinary ocean research, the long-standing flagship for physical and chemical oceanography.
• FORV Sagar Sampada (1984, CMLRE) - fishery and oceanographic research vessel; trawls, acoustics and plankton work supporting marine living resources assessment.
• ORV Sagar Nidhi (2008, NIOT) - ice-strengthened, dynamic positioning, deep-sea technology trials, mooring deployment, ROV support.
• RV Sindhu Sadhana and RV Sindhu Sankalp (CSIR-NIO, Goa) - coastal and open-ocean multidisciplinary cruises.
• Sagar Manjusha, Sagar Tara, Sagar Anveshika, Sagar Purvi/Paschimi - smaller coastal research and training vessels; INS Sandhayak class for naval hydrographic survey; a new generation of ocean research vessels has been contracted for the MoES fleet.
• Programmes: Deep Ocean Mission (polymetallic nodules in the Central Indian Ocean Basin, deep-sea mining technology, Samudrayaan); Indian Argo; Ocean Observation Network (moored buoys, OMNI/RAMA array in the Indian Ocean); Indian Scientific Expedition to Antarctica and the Southern Ocean (Maitri, Bharati stations); SIBER and Indian Ocean GOOS; Marine Ecosystem Dynamics of the Eastern Arabian Sea (CMLRE).
• Allied agencies: INCOIS (data, forecasts, potential fishing zone advisories), National Hydrographic Office Dehradun (charts), Fishery Survey of India (resource surveys), state pollution control boards and NIO for coastal EIA and monitoring.

Quality control and errors

• Calibration: CTD conductivity checked against bottle salinity measured on a laboratory salinometer; oxygen sensor against Winkler titration; fluorometer against extracted chlorophyll. The bottle data are the reference, the sensors give the resolution.
• Blanks, duplicates, standards: field blanks for nutrients and trace metals, duplicate samples (about 10%) and certified reference materials.
• Typical error sources: bottle misfire or leaking O-ring, salinity spikes from a fouled cell, net clogging (lowers filtration efficiency and underestimates abundance), wire angle more than 15-20° (net fishes shallower than wire-out suggests), grab closing on a pebble and washing out, contamination of trace-metal samples by the ship and the wire, and mislabelled bottles.
• Representativeness: a 5-litre Niskin bottle is a point sample of a patchy ocean; net hauls integrate but smear depth. Always report gear, mesh, volume filtered and depth range with any number.
• Metadata standards: time in UTC, position in decimal degrees (WGS84), SI units, depth in metres, salinity on the practical salinity scale (dimensionless, often written psu), and the gear and method named - without these the datum cannot be reused.

Key points

• Station - a fixed point where a defined measurement set is made; transect - a line of stations across a gradient.
• Stratified random sampling - divides the area into strata and randomises within each; gives an unbiased mean with a calculable error.
• CTD - measures conductivity, temperature and pressure; salinity is derived, not measured.
• Rosette - 12-24 Niskin bottles tripped electronically on the upcast around a CTD.
• Nansen bottle with reversing thermometer - the classic method; messenger-tripped, inverts to lock the temperature reading.
• XBT - expendable temperature probe from a steaming ship; depth from a fall-rate equation, no salinity.
• ADCP - current profiles from Doppler shift of backscattered sound; hull-mounted, lowered or moored.
• Secchi depth - transparency measure; euphotic depth is roughly 2.5-3 times Secchi depth (coefficient varies with water type).
• Draw order - oxygen first, then pH, nutrients, chlorophyll, salinity last.
• Winkler titration - the reference method for dissolved oxygen, used to calibrate oxygen sensors.
• Mesh size - 20 µm phytoplankton, 100-200 µm mesozooplankton, 330-500 µm larger zooplankton and fish larvae.
• Flowmeter - gives the volume of water filtered, without which a net haul cannot be made quantitative.
• Grab versus corer - grab takes a disturbed surface sample over a known area; corer preserves vertical stratification.
• Box corer and multicorer - undisturbed sediment-water interface for benthic ecology and geochemistry.
• Multibeam echosounder - full-coverage swath bathymetry; single-beam gives only a line of depths.
• Argo float - parks at about 1,000 m, profiles to about 2,000 m every 10 days; Indian data centre at INCOIS.
• Indian fleet - Sagar Kanya, Sagar Sampada, Sagar Nidhi, Sindhu Sadhana, Sindhu Sankalp, under MoES, NIOT, CMLRE and CSIR-NIO.
• UNCLOS Part XIII - marine scientific research in another state's EEZ needs that state's consent; India ratified UNCLOS in 1995.

Formulas and examples

Example 1 - zooplankton abundance from a vertical haul. A WP-2 net of mouth diameter 0.57 m is hauled vertically from 100 m to the surface. The sample is made up to 200 mL; a 5 mL aliquot contains 240 copepods.

• Mouth area = π × 0.57²/4 = π × 0.3249/4 = 0.2552 m².
• Theoretical volume filtered = 0.2552 × 100 = 25.52 m³.
• Individuals in the whole sample = 240 × (200/5) = 9,600.
• Abundance = 9,600 / 25.52 = 376 individuals m⁻³.
• If the flowmeter showed only 20.4 m³, filtration efficiency = 20.4/25.52 = 80%, and the corrected abundance = 9,600/20.4 = 471 individuals m⁻³ - ignoring the clogging underestimates abundance by about 20%.

Example 2 - benthic density from replicate grabs. Three Van Veen grabs of 0.1 m² each at one station yield 48, 61 and 35 polychaetes.

• Total area = 3 × 0.1 = 0.3 m²; total individuals = 144.
• Mean density = 144 / 0.3 = 480 individuals m⁻².
• Per-grab densities = 480, 610, 350 m⁻²; mean 480, deviations −0, +130, −130; sum of squares = 0 + 16,900 + 16,900 = 33,800; variance = 33,800/2 = 16,900; SD = 130 individuals m⁻².
• Standard error = 130/√3 = 75 individuals m⁻², so the station mean is 480 ± 75 - which is why one grab is never reported alone.

Example 3 - how many samples does the survey need? From the pilot above, s = 130 individuals m⁻², and the study wants the mean within ±50 individuals m⁻² at 95% confidence (t ≈ 2).

• n = (2 × 130 / 50)² = (260/50)² = 5.2² = 27.04 → 28 grabs.
• At 0.1 m² per grab that is 2.8 m² of seabed; tightening the target to ±25 would need n = (260/25)² = 108 grabs - precision costs quadratically, the single most useful fact in survey planning.

Example 4 - Secchi depth to euphotic depth and light. Secchi depth at a coastal station is 4.5 m.

• Kd ≈ 1.7/4.5 = 0.378 m⁻¹.
• Euphotic (1% surface light) depth = 4.6/0.378 = 12.2 m (≈ 2.7 × 4.5 = 12.2 m).
• Light at 5 m, with I₀ = 1,800 µmol m⁻² s⁻¹: I = 1,800 × e^{−0.378 × 5} = 1,800 × e^−1.89 = 1,800 × 0.1511 = 272 µmol m⁻² s⁻¹, about 15% of the surface value. The 1.7 coefficient is an approximation and varies from about 1.4 to 1.9 with water type, so quote it as such.

Example 5 - cruise time budget. A 20-station coastal grid, stations 25 nautical miles apart, ship speed 10 knots, each station needing a CTD-rosette cast (45 min), a plankton haul (30 min) and three grabs (45 min), plus 120 nm from port to the first station and back.

• Steaming between stations = 19 legs × 25 = 475 nm; port transits = 240 nm; total = 715 nm → 715/10 = 71.5 h.
• Station time = 20 × (45+30+45) min = 20 × 2 h = 40 h.
• Total = 111.5 h = 4.65 days; with a 25% weather and breakdown allowance, request 6 days of ship time. Under-requesting ship time is the commonest cruise-planning error.

Where it is used

• National research institutes: CSIR-NIO (Goa, with regional centres at Kochi, Mumbai, Visakhapatnam), NIOT (Chennai), CMLRE (Kochi), INCOIS (Hyderabad), NCPOR (Goa, polar and Southern Ocean cruises) - roles as project scientist, research associate, JRF, cruise participant and data manager.
• Fisheries and resource survey: Fishery Survey of India and ICAR-CMFRI use plankton nets, acoustics and trawl surveys to estimate larval abundance and stock; state fisheries departments use the same gear for coastal monitoring.
• Environmental consultancy and EIA: marine EIA for ports, SEZs, outfalls, desalination plants and offshore pipelines requires baseline water quality, sediment and benthos sampling under the EIA Notification, 2006 and CRZ Notification, 2019; NABET-accredited consultancies hire marine biologists specifically for this.
• Offshore oil, gas and submarine cables: ONGC and survey contractors run geophysical and geotechnical campaigns - multibeam, side-scan, sub-bottom profiling, vibrocoring - with marine mammal and environmental observers on board.
• Hydrography and defence: National Hydrographic Office and the Indian Navy's survey ships for charting; ports and dredging companies for pre- and post-dredge bathymetry and sediment testing.
• Deep Ocean Mission and blue-economy programmes: polymetallic nodule exploration in the Central Indian Ocean Basin, deep-sea biodiversity surveys, Samudrayaan submersible trials, ocean thermal and desalination technology testing at NIOT.

Interview questions

• Q: Why is salinity not measured directly by a CTD? A: The CTD measures electrical conductivity, temperature and pressure; salinity is computed from the three using the practical salinity scale algorithm, so a fouled or mis-calibrated conductivity cell shows up as a salinity error.
• Q: Why are rosette bottles fired on the upcast and not the downcast? A: The downcast profile must be undisturbed by the ship's wake and the frame's own turbulence; the bottles are then closed on the way up at depths selected from the live downcast profile.
• Q: In what order do you draw samples from a Niskin bottle, and why? A: Dissolved oxygen first, then pH/alkalinity, nutrients, chlorophyll, salinity last - gas samples are the ones degraded by exposure and bubbles, and salinity is the most robust.
• Q: What is a flowmeter on a plankton net for? A: It records the actual volume of water filtered so abundance can be expressed per cubic metre; without it, clogging and wire angle make the haul only qualitative.
• Q: Grab or corer - which would you use for benthic macrofauna and why? A: A grab (Van Veen, 0.1 m²) for routine macrofauna because it covers a known area quickly; a box corer where an undisturbed sediment-water interface or vertical structure is needed.
• Q: Why is formaldehyde for zooplankton always buffered? A: Unbuffered formalin becomes acidic and dissolves calcareous structures such as foraminiferan tests and crustacean exoskeletons, destroying the material for identification.
• Q: How do XBT and CTD differ? A: An XBT is expendable, launched from a moving ship, gives only temperature and infers depth from a fall-rate equation; a CTD is recovered, gives calibrated temperature, conductivity and pressure plus add-on sensors, but needs the ship to stop.
• Q: What does an ADCP measure and how? A: Current velocity in layers of the water column, from the Doppler frequency shift of sound scattered back by particles and plankton moving with the water.
• Q: Why do you still take Winkler titrations if the CTD carries an oxygen sensor? A: The sensor drifts and needs an absolute reference; discrete Winkler values at known depths are used to calibrate and correct the sensor profile.
• Q: A single grab shows 500 organisms m⁻². Can you report that as the station density? A: No - with one sample there is no variance and no standard error; at least three replicates are needed before a mean density can be quoted with confidence.
• Q: What permissions does a cruise in another country's EEZ need? A: Consent for marine scientific research from the coastal state under UNCLOS Part XIII, applied for well in advance, usually with data-sharing and observer conditions.
• Q: Your CTD profile shows a salinity spike at 60 m but temperature is smooth. What is the likely cause? A: A conductivity-cell artefact - trapped bubble, biofouling, or a sensor time-constant mismatch between the conductivity and temperature sensors - not a real water-mass feature; check against the bottle salinity.

Common mistakes

• Treating wire-out as depth: with current and ship drift the wire angle must be measured and corrected, or a pressure sensor used on the gear.
• Reporting plankton abundance from a net without a flowmeter, or ignoring filtration efficiency when the net has clogged.
• Drawing nutrient or chlorophyll samples before the dissolved oxygen sample, and allowing bubbles into the oxygen bottle.
• Preserving genetic material in formalin, or zooplankton in unbuffered formalin - both destroy the intended analysis.
• Single, unreplicated grabs or hauls per station, then quoting a mean density as if it had an error bar.
• Confusing a grab with a corer in the exam answer - a grab gives area-based density, a corer gives depth stratification.
• Omitting metadata (gear, mesh size, volume filtered, depth range, UTC time, position), which makes an otherwise good dataset unusable.
• Planning a cruise without weather contingency, and forgetting that some operations need daylight.

Related topics

Oceanography · Biological oceanography: plankton and primary productivity · Marine pollution monitoring and EIA · Benthic ecology and sediment analysis · Fisheries stock assessment · Marine instrumentation and remote sensing