AK SG 08 03l Video Analysis, Experimen

7/31/2019 AK SG 08 03l Video Analysis, Experimen

1/12

Marine Habitat Mapping Technology for Alaska, J.R. Reynolds and H.G. Greene (eds.) 157Alaska Sea Grant College Program, University of Alaska Fairbanks. doi:10.4027/mhmta.2008.11

AbstractSubmersibles are used in a diverse array o scientic stud-

ies such as geophysical mapping and interpretation, physical

and biological oceanographic studies, and shery investiga-

tions. Studies using submersibles deserve careul attention

to planning, design, and implementation in order to be eec-

tive in meeting their objectives. Te most important rst step

in these studies is to careully articulate and plan the exper-

imental design o the project rom start to nish. Specialattention should also be paid to statistical issues, such as ran-

domization, replication, independence, and power, as they

can have a large impact on the useulness o the data to meet

the objectives o the study. Te processing and analysis o vid-

eotapes resulting rom submersible dives involves a high level

o training and quality assurance with sta who are com-

ortable with taxonomic identication, survey methods, and

the wide array o technical tools used to analyze these data.

Issues associated with this video processing include train-

ing, minimizing observer bias, using the proper equipment,

and entering and veriying data. Large, complex projects

should always be managed using a relational database that

can eectively integrate complex data types, can validate data

types and ranges, and is signicantly less prone to errors than

spreadsheets. Given the high cost o conducting submersible

studies, it is imperative that sucient attention is ocused

on issues that impact data quality and that projects be well

thought out in their entirety prior to going to sea.

IntroductionScientic studies using submersibles are widely used to

investigate the abundance and distribution o shes, benthic

invertebrates, and their associated habitats. Te vehicles used

include human-occupied submersibles (HOVs), remotely

operated vehicles (ROVs), towed sleds, and more recently

autonomous underwater vehicles (AUVs). Projects involv-

ing observations rom submersibles may involve shery

biologists, marine ecologists, and marine geologists work-

ing in collaboration. Because these projects may integrate a

diverse array o studies including geophysical mapping and

interpretation, physical and biological oceanographic studies,

and shery investigations (e.g., Reynolds et al. 2001, Nasby

et al. 2002, Wakeeld et al. 2005), these projects typically

generate a large quantity o video data that require complex

and time-consuming post-processing and data management

(Somerton and Glendhill 2005).

Although there are numerous examples in the litera-

ture describing the results o such studies (e.g., Stein et al.

1992, OConnell et al. 2002, Yoklavich et al. 2000, Jagielo et

al. 2003, issot et al. 2007) there are ew published accounts

that capture the complex issues associated with the design

and implementation o these studies, especially post-pro-

cessing and data management issues. Tus, the goal o this

paper is to describe the design, implementation, and data

management o products typically resulting rom submers-ible studies. I will ocus on the technical and logistical issues

that cascade through these types o projects and how they

can be managed eectively and eciently both beore and

ater eldwork has been completed.

Experimental designBeore beginning any study the most important step is to

careully articulate and plan the experimental design o the

project rom start to nish (Green 1979). Detailed planning

at the start o the project can save signicant time, as later

changes in the design o the project will cascade through

multiple levels o the project such as eld surveys, video log-ging, database design, and geospatial and statistical analyses.

Initially, the objectives o the study need to be claried in

as much detail as possible. Based on the objectives, specic

hypotheses should be developed rom the outset in order to

clariy the statistical tests to be used, their assumptions, and

statistical power issues (Underwood 1991). Te eld sam-

pling strategy can then be derived rom the objectives and

hypotheses by taking into account the areas to be sampled,

time at sea, data collection capabilities, and other logistical

constraints (Krebs 1999). Once eld variables and the vari-

ous types o data to be collected are dened, data types and

structures can be specied or the relational database, which

can then be designed or data entry and validation.

Special attention should be paid to statistical issues rom

the outset as they can have a large impact on the useulness

o the data to meet the objectives o the study (Hurlbert 1984,

Eberhardt and Tomas 1991). Te principal issues include

randomization, replication, independence, and power. I

samples are large, random selection o study sites rom all

possible sampling locations should be used to minimize

sampling bias. Most sampling strategies use a stratied

design to ocus on particular depths and/or habitat types.

Appropriately sized grids are then overlaid on maps o the

study area and randomly selecting grids are used to select

Video Analysis, Experimental Design, and DatabaseManagement o Submersible-Based Habitat StudiesBrian N. Tissot

Washington State University, School of Earth andEnvironmental Science, Vancouver, Washington


2/12

158 TissotVideo Analysis, Experimental Design, and Database Management

sampling sites (e.g., Jagielo et al. 2003, Yoklavich et al. 2007)

(Fig. 1). I the number o possible grids is small relative to

those being sampled, it is better to systematically select sites

to be sampled to ensure appropriate interdispersion over the

study area (Hurlbert 1984). Replication should occur both

within and among strata (e.g., depth, habitat) to minimize

conounding o spatial variation. I the study area has signi-icant variation in habitat, the area should be stratied and

subsampled with nested replicates to urther account or spa-

tial variation (Underwood 1997). I replicates within grids

lack independence (i.e., transect segments within a dive)

they should either be pooled or treated as non-independent

repeated-measures. Finally, and importantly, the power o

the statistical design to reject the null hypotheses should

be examined i the study is ocused on developing baselines,

detecting change over time, or or control-impact studies

(Mapstone 1996, Krebs 1999). o address this important

design issue, power analyses should be conducted. Power

analysis requires some measure o the variability in the

measured traits, which can be derived rom pilot studies or

related studies, and the amount o change to be detected, or

the eect size (example in issot et al. 2007).

Survey preparationsPrior to eldwork it is imperative to provide training or eld

observers and to plan or the types o data to be collected

and their management, both at sea and in the lab. One o

the major advantages o HOVs over ROVs is that observers

in the sub can oten see, resolve, and record (via the audio

track) many more animals than are visible rom the video

cameras. However, training or submersible observers is

critically important and time consuming but essential to pro-

viding accurate counts o marine organisms. raining usually

involves a combination o pre-dive reviews o identication,

counts, and/or size estimates, underwater surveys super-

vised by an experienced observer (which could initially be by

scuba in some habitats), reviews o logged video with experi-enced observers, and/or comparisons to physical sh models

(e.g., plastic sh) o known size and abundance. In general,

observer bias in the band transect approaches, common

to submersible surveys, is less on conspicuous slow-mov-

ing taxa and greater on ast-moving and/or cryptic species

(Williams et al. 2006). However, with experience observer

bias tends to rapidly diminish (Williams et al. 2005) and di-

erences rom values recorded by well-trained individuals

can be quite small (Yoklavich and OConnell 2008).

During the cruise detailed notes should be taken o all

procedures, including lists o data types collected (e.g., tapes,

images, navigation les, samples), where the raw data les

are physically located, and the distribution o backup copies.Videotapes rom the submersible are a critical product o the

study and should be handled with special care. Ideally, two

copies o the raw video eed should be made so one tape can

be stored as a master and the other used to make additional

copies as needed. Not only is this procedure good practice

or protecting original data, but videotapes (analog or digi-

tal) can lose luminance ater a very ew plays. Te navigation

system (USBL, DVL, or other) needs to be time-synced with

the time code on the video, ideally by encoding on the audio

channel and/or the video overlay system. Any dierences in

the time codes o the various data sources need to be care-

ully checked and logged in order or data derived rom the

videotapes to be accurately linked to geospatial inormation(and thus habitat and bathymetric maps) and other inorma-

tion derived rom the submersible (e.g., oceanographic data,

depth, distance o bottom).

During the submersible surveys it is important to record

as much supportive inormation as possible or interpretation

o survey data and video back in the lab. Still photos, voucher

specimens, and comments recorded on the video audio chan-

nel all contribute to a better understanding o the various

products rom the study. Some towed sleds and ROVs, such

asROPOS(Shepherd and Wallace 2002), have surace oper-

ations that allow real-time rame grabs and/or data loggers

where ship-based observers can record observations, naviga-

tion inormation, etc. that can provide data and assist in video

interpretation later. Be sure to back up everything rom the

cruise such as videotapes, still photos, navigation, and bridge

logs, and distribute copies to colleagues or saekeeping. For

longevity, data should be archived with NOAAs National

Ocean Data Center (NODC) or other appropriate centers.

Video analysis and data loggingTe actual analysis and data logging rom the videotapes can

be very time consuming, so it is important to plan and bud-

get accordingly. Depending on the amount and complexity

Figure 1. Stratifed sampling design using depth (contours in m) andhabitat type (various colors) to randomly select samplinglocations or cowcod rockfsh (Sebastes levis) on Tanner andCortez Bank, Caliornia. A 1.5 1.5 km sampling grid is over-

laid on mixed sediment and rocky substrata areas at 75-300m depths. Dive tracks are indicated by green lines, whichwere subdivided by three 20 min transects (yellow symbols).Inset shows a typical dive track with transects. Based onYoklavich et al. (2007).


3/12

Marine Habitat Mapping Technology for Alaska 159

o the inormation to be extracted rom the tapes, analysis

time can exceed submersible time by as much as 5:1 or 10:1.

Tus, a cruise that collects 100 hours o videotape can take

500-1,000 hours or lab analysis (or 3-6 months ull-time

work or a single individual).

Te process o video analysis involves a high level o

training and quality assurance with sta who are comortablewith taxonomic identication, survey methods, and a wide

array o technical tools. Sta need to be amiliar with the var-

ious data types generated during eldwork (e.g., navigation,

oceanographic inormation) and the equipment used to ana-

lyze the data, such as video equipment, time code generator,

and relational database. I the same people are involved in

analysis they will also need to be amiliar with GIS, graph-

ing, and statistical analysis sotware.

For consistency, and to minimize observer bias, it is

essential to develop detailed protocols that describe each

o the tasks to be completed. Tese should be developed in

cooperation with all team members involved in the project

and modied as the project progresses and new situationsemerge. Protocols should describe the general procedure or

logging data, with specic examples embedded in the text.

Exceptions, caveats, and variations in the procedure due to

dierent submersibles should also be included. Once the pro-

tocol becomes stable it can be used to train new individuals

to conduct the work. An example o a procedure to classiy

physical habitats, based on the method in Stein et al. (1992),

is partially listed in able 1.

raining and quality assurance are essential or teams

o individuals working together on a project as the level o

human bias in classiying habitat and identiying, counting,

and sizing shes and invertebrates can be signicant. Ideally,

individuals directly involved in eld identication and count-ing should do the video analysis; otherwise new individuals

must be trained rom the beginning. raining to conduct

video logging should be done using a range o habitat types

and organisms likely to be encountered, and using tapes

that have previously been analyzed by experienced observ-

ers. Initially, the trainer should lead the trainee through the

process, ollowing the protocol and discussing issues as

they arise. rainees are then ree to log initial video on their

own. Ater a ew transects have been completed, the trainer

should review the video and trial data with the trainee, using

the protocol to discuss similarities and dierences between

the trial run and data collected by an experienced observer.

Te trial data can then be quantitatively compared to previ-

ous data or overall similarity in habitat coding, taxonomic

identication, counts, sizes, and other variables. Tis quality

assurance process, rom sample logging to quantitative anal-

ysis, should be repeated until the trainee can produce data

with a reasonable level o similarity to that rom an experi-

enced observer. In our lab 90% similarity is considered an

acceptable level o consistency and trainees can, or the most

part, then be let on their own to log data. However, or long-

term data quality, projects should be conducted within the

quality assurance cycle: plan the analysis, do the work using

the protocol, check periodically or data consistency, and act

to update the protocol and retrain individuals as needed.

Tere are a variety o ways to sta a video analysis lab-

oratory. In academic institutions where graduate students

are the primary source o labor, overlapping appointments

allows more experienced students to train newer ones, with

the aculty supervisor providing overall quality assurance.

Other models, also used in academic settings but more com-

mon in state and ederal agencies, are to use permanent sta

(technicians, biologists, or other) to conduct video analyses,

which ensures more long-term consistency, and/or to use

independent specialized consultants.When logging data rom videotapes it is important to

have a comortable workstation with easy access to all the

necessary equipment. At a minimum there should be a high-

quality video player (VCR or digital video player) with audio

(unless videotapes are digitized, see below), a high-resolution

monitor, and a computer. Data should be entered directly

into a relational database or validation. Other equipment

that can aid in the analysis might include key counters or

enumerating several dierent taxa simultaneously and a time

code generator, such as a Horita time code wedge (Fig. 2).

Depending on the complexity o the work, the workstations

can be occupied by a single observer, which is common or

sh observations and habitat classication, or 1-3 observers

dividing up taxa into manageable pieces, which is com-

mon or benthic invertebrates. Te video can be analyzed

in a single pass or using multiple viewings, with each run

ocusing on dierent taxa or making dierent observations

(e.g., size, counts, associations). In some set-ups, videotapes

rom multiple submersible cameras can be watched simul-

taneously to improve resolution and size measurements (e.g.,

Harvey et al. 2005). Video may also be randomly sampled to

shorten post-processing time. Subsampling o video usually

requires stratiying the sample by depth and/or habitat to

Figure 2. Typical video analysis laboratory set-up illustrating digitalvideo player, high-resolution monitor, counter, and com-puter with database.


4/12


Protocol or Continental Shel Habitat ClassifcationRevised: 10/18/2007

A. Recording habitat patches

1. Record the start and end time o each habitat patch (sections o contiguous habitat types). Start/end times corre-

spond to when the sizing lasers touch the next habitat type. I the lasers swing up and down onto and o o the new

patch, wait until the lasers stay on the patch to record the time. End time corresponds with the beginning o the next

unique habitat type.

2. ransitions to new habitat types are dicult to delineate when gradients occur; be consistent and move through

video several times to delineate changes; add a comment to the habitat entry i the delineation was dicult.

B. Identiying habitat types

1. Habitat patches must be at least 10 seconds long orDelta and 30 seconds long or ROPOSwith 10% error to allow

or play with the lasers. Tis means the absolute minimum patch length or Delta is 9 seconds, and or ROPOS is

27 seconds. Any patches o shorter duration should be excluded rom the analysis.

2. Patches must have two habitat codes, e.g., BC, MP. First code is 50 and


5/12


avoid biasing the sample and a power analysis or analysis o

species-areas curves to decide on the number o subsamples

to use (Krebs 1999).

Te video can be viewed directly rom the tapes or

accessed in digital orm on a computer. One advantage o

the latter is the ability to instantly locate any position on the

tape or analysis. A disadvantage is the time required to con-vert video to digital storage and the large amounts o hard

drive space needed to accommodate the les.

Habitat classifcation and visual surveyingOne commonly used method or classiying habitats, widely

used on the U.S. West Coast, is based on the procedure

described in Hixon et al. (1991) and Stein et al. (1992). For

examples and variations on this method see Anderson and

Yoklavich (2007), Love and Yoklavich (2007), Yoklavich et

al. (2000, 2007), issot et al. (2006, 2007), and Wakeeld

et al. (2005). Tis method classies physical habitats using

a combination o nine dierent categories o substrata andstandard geological denitions (e.g., Greene et al. 1999). In

order o increasing particle size or relie, these substrata are:

mud (code M), sand (S), gravel (G), pebble (P), cobble (C),

boulder (B), continuous fat rock (F), rock ridge (R), and pin-

nacles (). A two-character code is assigned each time a

distinct change in substratum type is noted in the video, thus

delineating habitat patches o uniorm type. Te rst char-

acter in the code represents the substratum that accounted

or at least 50% o the patch, and the second character rep-

resents the substratum accounting or at least 20% o the

patch (e.g., RM represents a patch with at least 50% cover

by rock ridge and at least 20% cover by mud). In some stud-

ies, habitat patches may also be assigned a code based on thedegree o three-dimensional structure as dened by the ver-

tical relie o the physical substrata relative to the seafoor

(issot et al. 2006, Love and Yoklavich 2007). Te area o

each habitat patch can be determined using navigation data

in ArcGIS to calculate habitat patch length and multiplying

by the width o the transect as determined empirically in

the eld or by delineated paths using sizing lasers. Te use

o lasers to delineate transect boundaries work best in low-

relie habitats. In more complex, high relie habitats they may

become problematic due to an increase in the chance o edge

eects associated with uncertainty in dening the boundar-

ies o the transect. Other methods to delineate transect width

include measuring the area o view with lasers xed at a setdistance (Yoklavich et al. 2000) and photogrammetric meth-

ods that grab rames o precisely dened area or analysis

(Jagielo 2004). Line transect methods (Buckland et al. 2001)

can also be used when transect widths are uncertain or when

the target species is uncommon (e.g., OConnell et al. 2002,

Yoklavich et al. 2007).

Fishes and benthic invertebrates are commonly sur-

veyed using methods adapted rom visual belt transects used

or scuba surveys on shallow rees (e.g., Sale and Sharp 1983).

HOV observations can be made through either orward or

starboard portholes, which are mirrored by corresponding

video cameras that overlap with the observers eld o view.

ROVs, towed sleds, and AUVs typically ace orward, down-

ward, or at an angle in relation to the substratum. ransect

areas o known width are commonly delineated by lasers or

other devices that allow the calculation o organism den-

sity. In HOVs, or ROVs where video is streamed real-time toscientists on the surace, observers can verbally tape-record

observations about the species, size or size class, abundance,

and behavior o individuals visible within the transect area.

For invertebrates, which are requently too numerous to

count, the ocus may oten be on individuals larger than 5

cm in size, or megaaunal invertebrates (issot et al. 2006).

Similarly, shes can be grouped and recorded into varying

taxonomic levels depending on the objectives o the study.

Once in the lab, video logging o data is acilitated

by reerence to audio logs o the dive, written notes, and/

or voucher specimens identied by taxonomic experts.

Additional observations may include microhabitat utilization

(Hart 2004), behavior (Punwai 2002), and/or associationsbetween sh, habitat, and structure-orming invertebrates

(Love and Yoklavich 2007; issot et al. 2006, 2007).

Database managementLarge, complex projects should always be managed using

a relational database, such as MS Access. In addition to

allowing multiple users to simultaneously access the data on

a network, a relational database eectively integrates com-

plex data types (included digital images and video), validates

data types and ranges, and is signicantly less prone to errors

than fat-le databases such as spreadsheets. Trough query-

ing, relational databases can allow seamless integration with

a wide variety o other programs including graphical and sta-

tistical programs, and with ArcGIS (Wright et al. 2007).

Relational databases can link several dierent kinds o

data using key elds in data tables. Data tables can include

cruise metadata, vessel and submersible data, sh and inver-

tebrate survey data, taxonomic inormation, habitat data,

and navigation and oceanographic inormation. Trough

reerential integrity, cross-reerencing o (or example) dive

numbers and taxonomic elds insures that valid inorma-

tion connects across multiple data sources. An example o a

database structure used to manage data rom a submersible

project is illustrated in Fig. 3.

Querying and data integrationOne o the major strengths o a relational database is the abil-

ity to manage large quantities o data and selectively integrate

and extract multiple types o data through database query-

ing. Unlike conventional fat-le databases, which require

sorting and/or rearranging data to provide data summaries,

relational databases do not change the structure o the under-

lying data but instead provide new data tables that represent

joins o existing data rom multiple sources (Fig. 4). Fish data,

or example, can be summarized by location, species, habitat


6/12


type, or time o day to make graphs or run statistical analyses.

By varying the level o querying (e.g., by dive, habitat patch,

depth, and/or time), data can be extracted at the proper level o

replication or complex statistical analyses, such as repeated-

measure ANOVA. Biological data can also be integrated with

data connected at multiple spatial and/or temporal scales such

as navigation, temperature, and depth, in order to create maps,

examine correlations, stratiy data, or any number o things

depending on the hypotheses being tested.

An additional major strength o relational databases isthe ease o integration o geographically reerenced data with

external applications such as ArcGIS and other spatial tools .

ArcGIS, or example, can directly read MS Access tables andcreate maps that integrate geospatial layers rom multiple

data types (Fig. 5). Tese types o maps can also generate spa-

tial data (e.g., nearest neighbor distances) that can be used in

statistical analyses to examine the randomness o biological

associations (Pirtle 2005, issot et al. 2006). Further, geospa-

tial tools such as the Kriging analysis can be used to examine

large-scale distributional patterns that can subsequentially

contribute to management plans (Pirtle 2005, Wakeeld et al.

2005). Kriging is a geospatial technique that uses least square

algorithms to interpolate the value o any data point at an

unobserved location rom observations o its value at nearby

locations. Tis method is illustrated in Fig. 6, which extrapo-

lates the results o a multivariate analysis that included data

on the abundance o sh, invertebrates and habitats derived

rom a submersible study at Cordell Bank, Caliornia.

Summary and conclusionsEective and ecient management o data generated rom

submersible studies clearly requires careul thought, plan-

ning, and training at multiple levels o the project. Te goal

o this paper is to illustrate important considerations that

must be made at various stages o the project, particularly

with respect to experimental design, video analysis and log-

ging, and database management. I proper attention is given

to the collection and management o data, and the training

o personnel involved in post-processing, studies can be con-

ducted that clearly meet their stated objectives and provide

Figure 3. Relationships among data tables in an MS Access relational database illustrating common felds and links among tables to establish

reerential integrity.


7/12


Figure 4. Examples o a query in MS Access that joins linked data tables(top) and the result o the query (let).


8/12


Figure 5. Map created with ArcGIS showing the distribution o the black coralAntipathes dendrochristos(green triangles) alongDelta submersible transects (line segments) in relation to species o rock-fsh observed, which are indicated by species codes (e.g., YOY, BANK). Depth contour shown in

meters. The geospatial link between fsh and invertebrate data was used to conduct a nearestneighbor analysis o fsh-invertebrate associations (Tissot et al. 2006).


9/12


Figure 6. Kriging analysis prediction map or Cordell Bank, Caliornia, produced using frst dimension mul-tivariate scores derived rom Correspondence Analysis o fsh, invertebrate, and habitat dataalong submersible transects (line segments). The map displays a spatial pattern rom shallow,

hard substrates (lighter shades) to deeper, unconsolidated substrates (darker shades) (romPirtle 2005).


10/12


relatively seamless integration among the spatial, graphical,

and statistical tools necessary to conduct the work. Given the

high cost o conducting submersible studies, it is imperative

that sucient attention is ocused on issues that impact data

quality and that projects be well thought-out in their entirety

prior to going to sea.

AcknowledgmentsI would especially like to thank my colleagues who are part

o the West Coast submersible research group and who

have directly or indirectly contributed to this paper: ara

Anderson, Mark Amend, Joe Bizzarro, Julie Clemons, Bob

Embley, Gary Greene, Mark Hixon, Milton Love, Bill Pearcy,

Susan Merle, Natalie Reed, Linda Snook, Rick Starr, David

Stein, Waldo Wakeeld, Curt Whitmire, and Mary Yoklavich.

My graduate students have been especially helpul and I have

learned more rom them than I have given: Camelia Bianchi,

Jen Blaine, Jennier Bright, Kaitlin Grai, Noelani Puniwai,

Jodi Pirtle, and Keri York. Te projects described in this

paper were unded by the West Coast and Polar Regions

Undersea Research Center o the NOAA National Undersea

Research Program, NOAAs Oce o Ocean Exploration,

NOAAs Oce o Habitat Conservation, NOAA Northwest

Fisheries Science Center and Southwest Fisheries Science

Center, NOAA Pacic Marine Environmental Laboratory,

and Washington State University. Id also like to acknowl-

edge all o the olks who operated the submersibles, ROVs,

and ships involved in the various projects during which these

methods were developed (Mermaid II,Delta,ROPOS, F/V

McGaw, NOAA shipRonald Brown, and F/VVelero IV). Te

paper beneted rom comments rom Jennier Reynolds and

three anonymous reviewers. North Pacic Research Board(NPRB) publication no. 171.

ReerencesAnderson, .J., and M.M. Yoklavich. . Multiscale habitat asso-

ciations o deepwater demersal shes o central Caliornia.

Fish. Bull. U.S. :-.

Buckland, S.., D.R. A nderson, K.P. Burn ham, J.L. Laake, D.L.

Borchers, and L. Tomas. . Introduction to distance sam-

pling: Estimating abundance o biological populations. Oxord

University Press, Inc., New York.

Eberhardt, L.L., and J.M. Tomas. . Designing environmental

eld studies. Ecol. Monogr. ():-.

Green, R.H. . Sampling design and statistical methods or envi-

ronmental biologists. Wiley, New York. pp.

Greene, H.G., M.M. Yoklavich, R.M Starr, V.M. OConnell, W.W.

Wakeeld, D.E. Sullivan, J.E. McRea, and G.M. Cai lliet. . A

classication scheme or deep seafoor habitats. Oceanol. Acta

:-.

Hart, . . Diel activity patterns in demersal shes on Heceta

Bank, Oregon. M.S. thesis, Oregon State University, Corvallis.

Harvey, E., J. Seager, M. Shortis, S. Robson, and M. Cappo. .

Use o stereo-video photography in remote camera and diver

transect assessments o sh populations. In: D.A. Somerton

and C.. Glendhill (eds.), Report o the National Marine

Fisheries Service Workshop on Under Water Video Analysis.

NOAA ech. Memo. NMFS-F/SPO-, pp. -.

Hixon, M.A., B.N. issot, and W.G. Pearcy. . Fish assemblages

o rocky banks o the Pacic northwest (Heceta, Coquil le, andDaisy banks). U.S. Minerals Management Service, OCS Study

-, Camaril lo, Caliornia.

Hurlbert, S. . Pseudoreplication and the design o ecological

eld experiments. Ecol. Monogr. ():-.

Jagielo, . . Remote Sensing using laser projection photo-

grammetry or underwater surveys. In: D.A. Somerton and

C.. Glendhill (eds.), Report o the National Marine Fisheries

Service Workshop on Under Water Video Analysis. NOAA

ech. Memo. NMFS-F/SPO-, pp. -.

Jagielo, ., A. Homann, J. agart, and M. Zimmermann. .

Demersal groundsh densities in trawlable and untrawlable

habitats o Washington: Implications or the estimation o

habitat bias in trawl surveys. Fish. Bull. U.S. ():-.

Krebs, C.J. . Ecological methodology, nd edn. Addison-Wesley

and Benjamin Cummings. pp.

Love, M.S., and M.M. Yoklavich. . Habitat characteristics o

juvenile cowcod, Sebastes levis (Scorpaenidae), in southern

Cali ornia. Envi ron. Biol. Fishes ./s---x.

Mapstone, B. . Scalable decision criteria or environmental

impact assessment: Eect size, ype I, and ype II errors.

In: R.J. Schmitt and C.W. Osenberg (eds.), Detecting ecolog-

ical impacts: Concepts and applications in coasta l habitats.

Academic Press, San Diego, pp. -.

Nasby, N., R. Embley, B.N. issot, M.A. Hixon, D. Wright, S. Merle,

and W.W. Wakeeld. . Integration o historical ground-

sh habitat data with new high resolution remotely sensed

data on Heceta Bank, Oregon. Fish. Bull. U.S. ():-.

OConnell, V.M., C. Brylinsky, and D. Carlile. . Demersal shel

rocksh stock assessment or . Alaska Department o

Fish and Game, Division o Commercial Fisheries, Regional

Inormation Report No. J-.

Pirtle, J. . Habitat-based assessment o megaaunal inverte-

brates and shes on Cordell Bank, Caliornia. M.S. thesis,

Washington State University. pp.

Puniwai, N.F. . Spatial and temporal distribution o the crinoid

Florometra serratissima on the Oregon continental shel. M.S.

thesis, Washington State University. pp.

Reynolds, J.R., R.C. Highsmith, B. Konar, C.G. Wheat, and D.Doudna. . Fisheries and sheries habitat investigations

using undersea technology. Marine echnology Society/IEEE

Oceans , Conerence Proceedings, MS ---.

Sale, P.F., and B.J. Sharp. . Correction or bias in visual transect

censuses o coral ree shes. Coral Rees :-.

Shepherd, K., and K. Wallace. . Deep precision deployment

and heavy package recoveries by theROPOSR.O.V. system.

Marine echnology Society, MS --, -.


11/12


Somerton, D.A., and C.. Glendhill (eds.). . Report o the

National Marine Fisheries Service Workshop on Under Water

Video Analysis. NOAA ech. Memo. NMFS-F/SPO-. pp.

Stein, D.L., B.N. issot, M.A. Hixon, and W. Barss. . Fish-

habitat associations on a deep ree at the edge o the Oregon

continental shel. Fish. Bull. U.S. :-.

issot, B.N., M.A. Hixon, and D. Stein. . Habitat-based sub-mersible assessment o macroinvertebrate and groundsh

assemblages at Heceta Bank, Oregon, rom to . J.

Exp. Mar. Biol. Ecol. :-.

issot, B.N., M.M. Yoklavich, M.S. Love, K. York, and M. Amend.

. Benthic invertebrates that orm habitat structures on

deep banks o southern Caliornia, with special reerence to

deep sea coral . Fish. Bul l. U.S. :-.

Underwood, A.J. . Te logic o ecological experiments: A case

history rom studies o the distribution o macro-algae on

rocky intertidal shores. J. Mar. Biol. Assoc. U.K. :-.

Underwood, A.J. . Experiments in ecology: Teir logica l design

and interpretation using analysis o variance. Cambridge

University Press.

Wakeeld, W.W., C.E. Whitmire, J.E.R. Clemons, and B.N. issot.

. Fish habitat studies: Combining high-resolution geo-

logical and biological data. In: P.W. Barnes and J.P. Tomas

(eds), Benthic habitats and the eects o shing. Am. Fish. Soc.

Symp. :-.

Williams, I.D., W.J. Walsh, B.N. issot, and L.E. Hallacher. .

Impact o observers experience level on counts o shes in

underwater visual surveys. Mar. Ecol. Prog. Ser. :-.

Wright, D.J., M.J. Blongewicz, P.N. Halpin, and J. Breman. .

Arc Marine: GIS or a blue planet. ESRI Press, Redlands,

Caliornia. pp.

Yoklavich, M.M., and V. OConnell. . wenty years o researchon demersal communities using the Delta submersible in the

northeast Pacic. In: J.R. Reynolds and H.G. Greene (eds.),

Marine habitat mapping technology or Alaska. Alaska Sea

Grant, University o Alaska Fairbanks. (Tis volume.)

Yoklavich, M.M., M.S. Love, and K.A. Forney. . A shery-inde-

pendent assessment o an overshed rocksh stock, cowcod

(Sebastes levis), using direct observations rom an occupied

submersible. Can. J. Fish. Aquat. Sci. :-.

Yoklavich, M.M., H.G. Greene, G.M. Cailliet, D.E. Sullivan, R.N.

Lea, and M.S. Love. . Habitat associations o deepwater

rockshes in a submarine canyon: An example o a natural

reuge. Fish. Bull. U.S. :-.


12/12

AK SG 08 03l Video Analysis, Experimen

Documents