In Situ Instrumentation

NeedS aNd ChalleNGeS Ocean-observing systems are changing the way ocean science is accomplished. No longer is ocean science limited to observations made by ships, whose scheduling and expense often constrain research to short forays that result in data streams limited in space and time. Such observations have been described as being “frozen in the invisible present,” offering thin slices of the ocean record that often miss processes that function on multiple spatial (e.g., boundary current, eddy, gyre, ocean basin) and tem-

3. How complex is the detection assay?
If simple staining or hybridization is required, results will be available in a relatively short period of time (therefore enabling a higher data-collection frequency), whereas amplification may take one to several hours.
4. Is it desirable to archive samples for examination and verification after instrument retrieval?Archiving requires some preservation as well as storage capacity of the system.
5. What are the design criteria for sensors in terms of size and power consumption?Size and power budget will limit the type of platform on which a particular sensor can be deployed (i.e., cabled observatory versus glider).(Lorenzen, 1966), and a wide variety of small, power-stingy sensors exist.
Variable fluorescence (Fv/Fm), based on saturation kinetics of Photosystem II, is used to determine key photosynthetic parameters for computation of phytoplankton primary productivity (Kolber and Falkowski, 1993).The current generation of variable fluorometers has typically been used in ship-based profiling or In-water spectrometers measure either a complete visible absorption spectrum or a limited number of wavelengths and have been used for months on moorings and days on mobile platforms.
In contrast to bulk optical properties, flow cytometers identify and count individual particles that stream past an array of light detectors.These instruments were originally intended for biomedical studies, but are now successfully used in the analysis of marine microbes (Chisholm et al., 1988;Olson et al., 1989;Binder et al., 1996;Shalapyonok et al., 1998).The use of flow cytometry is still largely restricted to the laboratory, but special instruments that can be deployed in the field are becoming available (Olson et al., 2003;Dubelaar et al., 1989; http://www.cytobuoy.com).Rapid advances in the technology, especially the use of solid-state lasers, will make it possible to deploy grids of flow-cytometry detectors at permanent observation sites.
Real-time, on-site plankton detectors will allow biological oceanographers to remotely observe the dynamics and spatial distribution of algae blooms and the proliferation associated microbes.

Molecular biological techniques
Although optical methods are highly evolved and used routinely in ocean science, they do not allow for distinction of many microbial groups, nor do they provide an indication of the genomic capacity (e.g., Culley et al., 2006;DeLong and Karl, 2005) Marcelino et al., 2006;Hashsham et al., 2004;Small et al., 2001).Using such methods in an autonomous system deployed in the ocean poses significant, though not insurmountable, challenges.studies that describe the temporal and geographical distribution of, most notably, the cyanobacteria (Chisholm et al., 1988;Legendre and Yentsch, 1989;Vaulot et al., 1995;Mann and Chisholm, 2000;Johnson et al., 2006).
Typical measurements of marine samples determine the forward scatter and side scatter and the fluorescence from chlorophyll and phycoerythrin (a reddish pigment found mainly in cyanobacteria and red algae).Chisholm and Vaulot and their collaborators have shown that these parameters are useful in measuring primary producers (Prochlorococcus and Synechococcus at different locations) (Vaulot et al., 1995).Li (1994) and Worden et al. (2004) use these parameters to quantitate picoeukaryotic grazers of the cyanobacteria.
Among the group of optical parameters that remains to be explored, only the use of scatter polarization has been reported.Olson et al. (1989) Monterey Bay Aquarium Research Institute, 2006. After Greenfield et al. (2006) has automated application of three different classes of DNA probe arrays in single field deployments lasting 20 days, targeting detection of marine planktonic organisms ranging from heterotrophic and photosynthetic bacteria, archaea, and harmful algae to small invertebrates found in the upper ocean (Christina Preston, Monterey Bay Aquarium Research Institute, pers. comm., 2006;Goffredi et al., 2006;Greenfield et al., 2006;Babin et al., 2005).A competitive ELISA (Enzyme-Linked ImmunoSorbent Assay) for the algal biotoxin domoic acid, a neurotoxic amino acid, was also fielded in concert with the probe arrays (Gregory Doucette, NOAA/National Ocean Service, pers. comm., 2006; http://www.mbari.org/microbial/esp/esp_technology.htm).This is the first record of sensing in situ both a harmful algal species and the toxin it produces (an amino acid metabolite) using molecular probe assays.(Ottesen et al., 2006).
No doubt much work remains to define the assays that will be deployed in , ChrIS SCholIN, Ger VaN deN eNGh, aNd Mary JaNe Perry NeedS aNd ChalleNGe S Ocean-observing systems are changing the way ocean science is accomplished.No longer is ocean science limited to observations made by ships, whose scheduling and expense often constrain research to short forays that result in data streams limited in space and time.Such observations have been described as being "frozen in the invisible present," offering thin slices of the ocean record that often miss processes that function on multiple spatial (e.g., boundary current, eddy, gyre, ocean basin) and temporal (e.g., monthly, seasonal, annual, decadal) scales.The key to autonomous observations of microbes in the ocean is continuing development of sensing technologies in the laboratory, transitioning sensors from the bench to the field, and integrating sensor suites into observing platforms appropriate to the spatial and temporal dimensions of specific processes and phenomena.With regard to platforms, the last sev-eral decades have witnessed an impressive evolution of in-water platforms that extend the temporal and spatial reach of ships.Bottom-tethered coastal and deepsea moorings provide time-series data at single locations (i.e., OASIS: http:// www.mbari.org/oasis)and as integrated observing networks (i.e., GoMOOS: http://www.gomoos.org).Enhanced battery life and new technologies that locally produce energy are enabling longer mooring deployments and additional instrumentation.More recently, the development of shore-powered, cabled observatories with high bandwidth is freeing researchers from constraints of power limitation and enabling rapid two-way communication with sensors and other devices (i.e., Martha's Vineyard Cabled Observatory: http://www.whoi.edu/mvco/description/description2.html; Venus: http://www.venus.uvic.ca;LEO-15: http://marine.rutgers.edu/cool/LEO/LEO15.html;and others in planning).Mobile platforms such as profiling floats, drifters, autonomous underwater vehicles (AUVs), and gliders allow questions to be addressed on a range of spatial scales; mobile platforms either follow water masses in a Lagrangian mode or operate in a survey mode (Rudnick and Perry, 2003).Distributed networks of diverse and complementary oceanobserving systems offer the possibility of integrated, continuous, real-time observing of oceanic phenomena over large areas without the limitations imposed by shipborne observations (Figure 1).Despite the successes of moorings, gliders, and other observational platforms in routinely making long-term autonomous measurements of physical or meteorological data, biological sensing systems-particularly those capable of microbiological measurements-are in their infancy.With a few notable exceptions, most autonomous biological sensing systems are optically based and JohN Paul (jpaul@marine.usf.edu) is Distinguished University Professor, College of Marine Science, University of South Florida, St. Petersburg, FL, USA.ChrIS SCholIN is Senior Scientist, Monterey Bay Aquarium Research Institute, Moss Landing, CA, USA.Ger VaN deN eNGh is Research Professor, University of Washington, Seattle, WA, USA.Mary JaNe Perry is Professor, University of Maine, Walpole, ME, USA.> SeCtIoN III.toolS, MethodoloGIeS, INStruMeNtatIoN, aNd aPProaCheS > ChaPter 6. IN SItu INStruMeNtatIoN ...biological sensing systems-particularly those capable of microbiological measurements-are in their infancy.a S e a o f M I C r o b e S typically focus on bulk optical measurements.In contrast, laboratory-based technologies include rapidly evolving, highly capable molecular techniques for taxonomic and functional analysis and optical methods for analysis of single cells.The challenge for observatories is to transition technology capais the concentration or frequency of occurrence of the target organisms?Certain targets may always be present at a relatively high concentration (i.e., bacteria) while others may only occur episodically (i.e., harmful algae), and yet others (human pathogens) may be so dilute as to require sample sizes in the hundreds of liters.2. What is required for sample preparation prior to analysis?Certain detection technologies require nucleic acid extraction and purification, while others require staining or probe hybridization to nearly intact cells.Simple sample preparation is certainly better than a lengthy series of extraction and purification steps.
6. How long can the sensing system (sensor and platform) operate between service visits?Biofouling, stability of reagents, and sample capacity are among the factors that will determine frequency of sampling and length of deployment.Ultimately, a desirable goal is service frequencies of months (even better, years).aPProaChe S optical techniques Optical methods have long been used to study autotrophic phytoplankton, either at the community level or as individual cells.Chlorophyll a fluorescence is widely used to assess phytoplankton abundance

figure 1 .
figure 1. Vision of the components of an ocean-observing system, including cabled observatories, autonomous underwater vehicles, gliders, buoys, moorings, satellites, and a traditional observing platform (research vessel).Image courtesy of Harris Maritime Communications For example, STMicroelectronics offers the In-Check® platform, a microfluidic chip that combines PCR amplification and probe array detection functions.Integrated devices like this system could find application for deploying "conventional probe array chemistries" in an ocean setting.criteria for the identification of microbes by flow cytometry and have conducted extensive field observed that differences in polarization of forward scatter can be used to distinguish among coccolithophores, diatoms, and other microbes.Scatter depolarization is a promising parameter to determine the degree of calcification of coccolithophores, and it may be useful in determining the productivity and carbon fixation of this ecologically important group of microorganisms(Iglesias-Rodriguez et al., 2002, 2006).Recent engineering efforts by author van den Engh and Tim Petersen, now at Cytopeia, Seattle, have led to greatly improved detectors for polarized scatter measurement.This new generation of detectors can register particles as small as 100 nm and determine scatter-and fluorescence-depolarization with great precision.When combined with photomultipliers with a high current capacity, the dynamic range can be adjusted to cover six or even eight decades of signal intensity.Flow cytometers are complex instruments, and their fragile character is an obstacle for use in the field.Historically, flow cytometers used finicky, powerhungry lasers.This situation is rapidly changing.In recent years, a wide range of solid-state lasers has become available.At this moment, solid-state lasers offer a wide choice of wavelengths and light intensities between 355 nm and 700 nm.The availability of adequate light sources no longer is an obstacle to field applications.Current flow cytometers require a particle-free carrier fluid to transport particles through the measurement area.Prolonged operation at a remote location requires a constant supply of clean sheath fluid.The two systems that have been built for use at sea recycle the carrier fluid and remove particles by filtration as new sample is injected into the core of the fluid stream.The mechanism that Rob Olson and Heidi Sosik (Olson et al., 2003) developed for their system is remarkably robust and has operated for months at the test site.A plankton detector that does not require a sheath fluid is being developed (Jarred Swalwell, School of Oceanography, University of Washington, pers.comm., 2006).The detection system of this instrument determines the position of the particles in front of the detector (Position Sensitive Detector, PSD).Only particles that follow a trajectory through the optical optimum are accepted for analysis.The PSD has been shown to perform accurate measurements on unfiltered seawater flowing though a simple fluidic system.Developments like this will lead to simpler designs with increased reliability and longevity in the field.optical Phytoplankton discriminator The Optical Phytoplankton Discriminator (OPD) (Figure 2) is a highly adaptive phytoplankton-sensing module developed by Mote Marine Laboratory, Sarasota, Florida, under the direction of Gary Kirkpatrick (Robbins et al., 2006).The instrument is designed to discrimi-nate the Gulf of Mexico red tide organism Karenia brevis from other phytoplankton based upon optical properties.The heart of the module is a liquid waveguide capillary cell (LWCC) attached to a fiber-optic spectrometer, illuminated by a fiber-optic tungsten/deuterium light source.The operational sequence of this instrument is to first draw a sample into the LWCC, take a spectral reading, and then draw in a reference solution from an onboard reservoir to take a reference spectrum.Finally, the LWCC pulls in a filtered (cell-free) sample of the ambient water to get the spectral properties of the dissolved components of the sample in question.Pigment absorbance peaks are transformed using fourth derivative analysis and compared to values obtained with a reference K. brevis culture.A similarity index is computed that ranges from 0 to 1, a value of 1 being most similar to K. brevis.The OPD can be deployed on stationary moorings or mobile platforms such as the BSOP (Bottom Stationed Ocean Profiler; http://cot.marine.usf.edu/Bsop/Bsop.htm) and autonomous underwater vehicles such as gliders and REMUS (Remote Environmental Monitoring UnitS) (Robbins et al., 2006).A distinct advantage of the OPD is the minimal sample preparation time that enables it to process multiple samples quickly, as required for AUV deployment.

Figure 3
Figure 3 shows data obtained from the deployment of the OPD on an AUV off the coast of southwestern Florida in January 2005.The proportion of the phytoplankton attributed to K. brevis is reported in conjunction with salinity (reported as density).These data show that K. brevis is more abundant in the western portion of the transect (left side of figure).

figure 3 .
figure 3. Cross section of water density and Karenia brevis chlorophyll biomass fraction obtained from a brevebusterequipped glider on January 15-16, 2004.from the beginning of the plot to approximately 2130 hrs on January 15, the glider was moving west-southwest across the shelf.It then turned and proceeded southeast, parallel to the coast, until it was recovered.due to the sampling scheme of the brevebuster, the vertical positions of the biomass fraction values are rough approximations.These positions could vary by approximately 50% of the bottom depth.although it is not possible to give the depth of the K. brevis observations precisely, the horizontal distribution shows a higher biomass fraction at the northern (left side) extent of the survey.Note that the density values are individual measurements, not contours.Gary Kirkpatrick, Mote Marine Laboratory

figure 4 .
figure 4.These are 16s rrNa-targeted dNa probe arrays printed with probes for marine microbial groups developed using the eSP supplied with different samples.The bottom panel shows the pattern of probes and an abbreviation of the group targeted.The top panel shows the actual arrays exposed, left to right, to a lysis buffer only, a sample collected near the surface, and a sample collected at 200 m.The arrays demonstrate change in the microbial community as a function of depth, quantified as mean pixel intensity in the middle panel.The actual size of the arrays shown are ~ 15 mm 2 .Figure courtesy of Christina Preston,Monterey Bay Aquarium Research Institute, 2006.After Greenfield et al. (2006) figure 5.The second-generation environmental Sample Processor (2G eSP) being tested in a seawater tank ahead of deployment in Monterey bay.The instrument is moored subsurface and an electromechanical cable provides for communications between a remote station and the eSP's surface buoy.an integral conductivity-temperaturedepth (Ctd) package is visible at left.The eSP operates on 12-volt rechargeable batteries (at bottom, above the anchor).Photo credit: Todd Walsh, Monterey Bay Aquarium Research Institute situ and the concomitant, upstream sample collection and processing require-ments.Putting all the pieces together from a systems point of view remains a ripe area for future investigation.Sustained investment in the development of small, robust, in situ instrumentation is essential to bring to fruition the testing of ideas and models discussed in this special issue.aCkNowled GeMeNtS This work has been supported in part by grants from ONR, NSF, and NOAA-ECOHAB to J.H.P., from NSF 0451010 and 0526231 to M.J.P. and from NSF 0314222, and the Gordon and Betty Moore Foundation ERG 731 to C.A.S. refereNCe S Ahn, S., D.M. Kulis, D.L. Erdner, D.M. Anderson, and D.R. Walt.2006.Fiber-optic microarray for simultaneous detection of multiple harmful algal bloom species.Applied and Environmental Microbiology 72:5,742-5,749.