Introduction

The phenomenon of Epizootic Shell Disease (ESD) arguably destroyed commercial lobstering south of Cape Cod. Its incidence has increased north of Cape Cod in the Gulf of Maine and Georges Bank and we can ask whether it is following a similar pattern in the Gulf of Maine where it could be a major economic factor in the region for both Maine, New Hampshire, Massachusetts as well as Canada. The NOAA NE Groundfish Survey, paused in Spring 2000 due to the pandemic, was re-established in Spring 2021 and the protocol developed for surveying for ESD was carried out.

The relevant data sets from NOAA NMFS were transformed to provide appropriate input for analysis, testing the assumption of random occurence of ESD. If uniform across the Gulf of Maine and Georges Bank, it could be evaluated by a Chi Square test of the contingency tables of numbers of ESD and non-ESD lobsters observed in trawls, clustered so as to make numbers of ESD lobsters high enough within groups to allow clusters to be observed and tested if they meet the low sample number limits on testing contingency tables using Chi-Square methodality (McHugh 2013). If clustered in the available data, one could proceed to explore what was causing the clustering.

This report focuses on one season of data, Spring 2021, but could be applied retrospectively to the other three available data sets Fall 2018, Spring 2019 and Fall 2019. Application of the Chi-Square test to sequential strings of trawls from the Gulf of Maine and Georges Bank in Spring 2021 suggests ESD is clustered and Monte Carlo examinations of random collections of trawls confirm the importance of the sequential sets as capturing that clustering. This clustering has relevance in the further discussion of whether ESD should be more correctly considered enzootic or epizootic which is also related to discussion of 'areas of endemism' (Cassagranda and Lizarralde de Grosso 2013).

Methods

Data Collection

The collection of American lobster ESD data was accomplished using the protocols established in the FSCS2.0 computer directed protocol that involves protocols for all the NE Groundfish Survey organisms and includes online guides at four workstations devoted to the survey. Three of the workstations are manned each by a team of two, a cutter and a data recorder. The FSCS2.0 computer directs the protocol to be followed for each organism being worked on at a given workstation. A head workstation is occupied by the watch chief who decides how many baskets of an organism will be processed in each trawl. If a very large catch of an organism has occurred at a trawl site the watch chief will do a sub-sample of the baskets to establish which organisms are to be worked up at a team workstation. Chosen samples are then worked up and data on lobsters includes the trawl number that has metadata including all physical information about the trawl site (time, latitude, longitude, depth, and CTD (concentration, temperature, density) associated data). Within the sample each lobster has its sex, carapace length (CL), condition of legs and claws, presence of barnacles determined and state of ESD determined on a 4 point scale (0-3) (Smolowitz et al. 2005). ESD lobsters graded 1-3 were frozen for validation and measurements on land.

Statistics

The applicability of the Chi-Square method was based on numeric limits described by McHugh (2013).

Results

Preliminary results from Spring 2021 NOAA NMFS are available at URL:
http://www.bio.umass.edu/biology/kunkel/pub/lobster/BigS2021/S2021_ESDsurvey.html

Bibliography

1. Dolores Casagranda and Mercedes Lizarralde de Grosso (2013). Areas of Endemism: Methodological and Applied Biogeographic Contributions from South America. Chapter 1. InTech
2. Mary L. McHugh (2013). Lessons in biostatistics: The Chi-square test of independence. Biochemia Medica 23(2): 143-9.
3. Joseph G. Kunkel (March 27 2021) Is shell disease in the American lobster epizootic or enzootic.
4. Kathleen Reardon, Robert Russell, Rebecca Peters. (2020). 2020 Lobster Monitoring Update.
5. Smolowitz RM, Chistoserdov AY, Hsu A (2005) A description of the pathology of epizootic shell disease in the American lobster, Homarus americanus H. Milne Edwards 1837. J Shellfish Res 24:749−756
6. Joseph G. Kunkel (February 21, 2021) Landmark analysis of Homarus americanus including outlines of esd lesions etc.
7. Joseph G. Kunkel (Jan 16, 2021) Lobster Shell Disease in NOAA Northeast Groundfish Surveys 2018-2019
8. Joseph G. Kunkel (February 16, 2019) The American lobster Homarus americanus normal berried adult female.
9. Joseph G. Kunkel (November 20, 2018) Training website for recognizing Epizootic Shell Disease, ESD

At the 2021 National Shellfish Association annual meeting, NSA2021, which was presented on Zoom and discussed either on Zoom or Discord, I had presented a report on shell disease in Homarus americanus, which I called Epizootic Shell Disease (ESD). It was enlarged upon in a iNat Report on my Shell Disease Project submitted a month earlier (Kunkel 2021). During the post talk discussion, Roxanne Smolowitz asked me the probing question of why I called it ESD rather than Enzootic Shell Disease? I basically could not answer based on current data I have on hand but argue here that I had, perhaps erroneously, assumed that the disease had shifted north with the center of the lobster population. Colleague Brian Tarbox and I had been following a population of what we thought was a hotspot of 'ESD' from traps in the waters just outside of Casco Bay while traps inside Casco bay were not seeing any 'ESD'. I will assemble the data here that would be a basis of deciding the Epizootic vs Enzootic label and invite discussion from anyone interested. There remains and issue of why it was called Epizootic south of Cape Cod and why it should not be called Epizootic as well north of Cape Cod and particularly in the Gulf of Maine and Georges Bank areas which I have been surveying with the NOAA NE Groundfish Survey in the generally offshore population resident at greater than 50 meters.

Summary

Body

References

Kunkel JG (2021). Landmark Analysis of Homarus americanus including outlines of ESD lesions, etc. URL

Meres NJ, CC Ajuzie, M Sikaroodi, M Vemupalli, JD Shields & PM Gillvet. (2012). DYSBIOSIS IN EPIZOOTIC SHELL DISEASE OF THE AMERICAN LOBSTER (HOMARUS AMERICANUS). J Shellfish Research 31(2) 463–472.

Malloy SC. (1978). Bacteria induced shell disease of lobsters (Homarus americanus). J. Wildl. Dis. 14:2–10.

Smolowitz R, AY Chistoserdov and A Hsu. (2005). A DESCRIPTION OF THE PATHOLOGY OF EPIZOOTIC SHELL DISEASE IN THE AMERICAN LOBSTER, HOMARUS AMERICANUS, H. MILNE EDWARDS 1837. J Shellfish Research, 24(3) : 749-756 Download

Introduction

Part of my interest in American lobsters is their shapes and how Epizootic Shell Disease (ESD) lesions fit into their shape. Early on we in the lobster shell disease community observed that the lesions often centered on the dorsal of the carapace where we noticed that grooming using their leg claws was not reachable. This was also confounded with the development of an approach to estimating the coverage of the carapace with ESD on a 0-3 point scale of ESD severity developed by Bruce Estrella (Smolowitz et al. 2005).

Shell disease index used to assess the severity and proportion of lobsters affected.*

0	=	no shell disease symptoms
1	=	shell disease symptoms on 1–10% of the shell surface
2	=	shell disease symptoms on 11–50% of the shell surface
3	=	shell disease symptoms on >50% of the shell surface
	*	Presented by B. Estrella to the Lobster Shell Disease Workshop, Connecticut Sea Grant College Program, Waterford, CT, June 15, 2000.

Given that the estimation of ESD coverage was to be a practical scale to apply perhaps at sea during a sorting operation it was thought we should validate the approach quantitatively based on how the sorting was done. Here we present an approach to quantify the visual scale on a group of lobsters for which we were taking landmarks of the carapace.

Methodology

Landmark analysis is described for two dimensions by Rohlf and Slice (1990) and generalized to three dimensions. A 3-D digitizer was used to collect the x,y,z-locations of twelve landmarks that were known to be homologous in all individuals unless obscured by an encrustation or a lesion. A Microscribe G2X digitizer was used to manually collect the locations of the landmarks in a lobster that was immobilized on a metal breadboard with convenient posts that could be adjusted for the lobster size, Fig 1.

Fig. 1. Emily Whalen, a Teacher at Sea on NOAA Ship H.B. Bigelow, collects landmarks from a lobster immobilized in a breadboard with posts using bungi cords. The digitizer is driven by a PC running Microscribe proprietory software.

Two sets of landmarks were collected for each carapace sequentially so as to determine the reproducibility of the digitizer. A linear set of data which corresponds to the rout of the dorsal suture was recorded and plots in blue, additional linear records of lesion circumferences were recorded as appropriate fig 2.

Fig.2. Landmarks and linear features collected by Microscribe G2X digitizer, data saved to a CSV file and plotted with an R standard library of plotting functions.

The three dimensional surface of the carapace of the lobster was modeled using photogrammetry which takes several images of the surface of an object and stitches them together using the software Metashape ver. 1.6.5 (Semyonov 2018) to produce a three dimensional surface that can be overlain with the color and texture of the surface, looked at from any chosen angle and saved as an .obj file or save with various sized rotating videos.

Each carapace, placed vertically on a rotatable base, is rotated manually by approximately 10 degrees in 10 to 12 steps and digital images of 2000 x 3008 pixels are taken with a tripod mounted digital camera (Nikon D40 w. 18-55mm Nikkor lens). The images, fig 3, are submitted to Metashape and merged into a 3-dimensional surface model masking out the non-carapace structures. Results are stored as an obj. file, can be examined further with Metashape, or separately with Meshlab (Berejnov 2009) to point to structures such as lesions or encrustations such as barnacles for quantification. Lesions can also be quantified by identifying their center and circumference using the Microscribe G2X digitizer, but that must be applied on the specimen itself while the Meshlab approach provides a method for quantification using the 3-D model stored as an .obj file. The meta-data on location, depth, temperature, date/time can then be stored with an emended .obj file in a database.

Fig. 3. Modeling Lobster Carapace Surface, Color and Texture. A. Raw Images of a lobster from sequential rotational angles at three axial levels. ... 35 images that are the basis for computing Models: B. Link to SketchFab rotateable carapace model. C. Link to AVI of horizontally rotating carapace model.

A. Ha161 1	Ha161 3	Ha161 5	Ha161 7	Ha161 9	Ha161 11	See Models
						B. SketchFab 3D Ha161-01 model C. AVI Video of Model
	-

The three dimensional model .obj file can be loaded by a variety of secondary software, oriented and viewed optimally in a computer screen using the software Meshlab in which secondary notations can be added for landmarks, encrustations and lesions, fig 4.

Fig.4. Landmarks and other features such as ESD lesions or encrustations collected and labeled using Meshlab, data saved to the .obj file and recallable and replottable with labels using Meshlab. Here we see landmarks and lesions marked and labeled within the Meshlab work space. An option is available to save the current scene.

The .obj file can be presented on the internet using the website Sketchfab allowing manual rotation and magnification by the viewer.

Results

Distribution of lobsters, lobster shape and ESD in the Northeast Groundfish Survey area concentrating on the Features of the Gulf of Maine and Georges Bank.

ESD Lesion Development based on Photogrammetry Models

Construction of 3-dimensional models of the lobster carapace cuticle surface using photogrammetry allows the resolved surface features of the cuticle to be studied from any direction after the lobster is no longer available. I present one example of that approach by analysis of ESD lesions in a model of a lobster with level 2 ESD caught in Spring 2019 trawl 398. A model of lobster 1 from that trawl was created that can be viewed in a link to Sketchfab. That model was loaded into Meshlab and rotated to focus on four aspects of ESD lesion development. Fig 5A the symmetry of ESD lesions. Fig5B the origin of ESD lesions. Fig5C the merger of ESD lesions. Fig 5D, the development of ESD lesions.

Fig.5. Model Use of Spring 2019 lobster 398_01. A. Symmetry of lesions. B. Origin of lesions. C. Development of lesions. D. Merger of lesions. You can view a larger version of the image using keyboard control language. {cntrl-click on image and choose view}

A.		B. C. D.

Consequences for Carapace Cuticle Model

The consequences of the sequential development of lesions can be viewed in Model 16, fig 6, of the Lobster Carapace Cuticle in the progression through the four defined layers of the cuticle starting with penetration through the apli-cuticle coded brown, then penetration through the epicuticle which is mainly protein and calcite devoid of chitin. It is also devoid of prophenoloxidase and thus does not tan dark brown. Next the lesion penetrates into the exocuticle which is layered with chitin and protein including pro-phenoloxidase which when activated creates a dark brown melanization. Lastly the lesion penetrates the endocuticle.

Fig.6. Carapace Cuticle Model 16. ESD lesions sequentially penetrates through the apli-cuticle, the epicuticle, the exocuticle and the endocuticle. Based on published model and data (Kunkel et al. 2018).

Distribution of ESD in Lobster Populations

Preliminary results on the distribution of ESD within the NOAA Northeast Groundfish Survey for years 2018 and 2019 were assembled using scientists on legs 1-4 of the survey. The lobsters were examined from each trawl that yielded lobsters and either all lobsters were examined or a subsample was examined. The examined lobsters were measured for their carapace length, sex, presence of eggs, presence of barnacles, ESD score (0-3). If graded ESD 1-3 the lobster was frozen for transport to shore and confirmation of ESD and Photogrammetry. Fig 7 illustrates the counts of ESD and total lobsters at locations in the areas of the Groundfish Survey that yielded lobsters.

Fig.7. Distribution of Lobsters and ESD in the Northeast Groundfish Surveys of 2018 and 2019 with respect to historical contour depths.

Shapes of Lobster Populations

Six populations of lobsters were identified based on geological location around the Gulf of Maine. The center of the Gulf of Maine is relatively empty of lobsters as seen in fig 7. Based on this geography we can define Georges Bank, Scotian Shelf, Off-shore Coastal Maine, and Cape Cod Bay populations that were collected from trawls of the NE Groundfish Survey. In addition an Inshore Coastal Maine population was sampled from the inshore lobster fishery. The average shapes by Landmark Analysis of these populations were determined and one can see the departures of each population from the average structure In figure 8.

Fig.8 Shape Vector Departures of Lobster populations from the average in the Gulf of Maine. A. Static vector departures. B. Rotating 3-D vector departures. C. Zone 2-D shape grid departures.

A.	B. Stereo AVI C. Shape Deformation Grid Table

Those vectoral differences in shape can be interpreted as relative similarities and used to calculate a distance matrix Table I. Scotian Shoales and Georges Bank populations were decided to be too close to discriminate and were merged.

Table I. Distance between the average shapes of males (m) and females (f) of four lobster populations around the Gulf of Maine.

Grp1m Grp1f Grp2m Grp2f Grp3m Grp3f Grp4m Grp4f

Grp1m 0.000 0.07311095 0.1564157 0.1731994 0.11275326 0.14253301 0.08539479 0.12330810

Grp1f 0.07311095 0.000 0.1483276 0.1601467 0.09856017 0.10538918 0.07270127 0.07505296

Grp2m 0.15641573 0.14832755 0.000 0.1101980 0.11614984 0.09869060 0.13277203 0.13320282

Grp2f 0.17319941 0.16014672 0.1101980 0.000 0.13618227 0.10229913 0.15301850 0.14416986

Grp3m 0.11275326 0.09856017 0.1161498 0.1361823 0.000 0.07940803 0.13277203 0.06216734

Grp3f 0.14253301 0.10538918 0.0986906 0.1022991 0.07940803 0.000 0.24801883 0.14416986

Grp4m 0.08539479 0.07270127 0.1327720 0.1530185 0.13277203 0.24801883 0.000 0.04918337

Grp4f 0.12330810 0.07505296 0.1332028 0.1441699 0.06216734 0.14416986 0.04918337 0.000



The distance matrix is used to calculate a shape similarity tree, fig 9

Fig.9. Shape Similarity Tree of Lobsters from the Gulf of Maine calculated on the distance matrix of Table I using the R library .

Discussion

The shape of the American lobster does differ between populations distributed around the Gulf of Maine. However the distribution of ESD seems to be somewhat even over the entire Gulf of Maine and Georges Bank as seen in the NOAA 2018-2019 NE Groundfish Survey seasons. One feature of this recorded ESD and normal data set is that the NE Groundfish Survey reports mainly on depths greater than 50 meters ignoring much of the inshore lobster population. The only segment of the groundfish survey that covers the inshore lobster population is in Cape Cod Bay. In that respect the shape relationship tree shows Cape Cod Bay carapace shapes to be closest to Inshore Main Coast lobsters. The carapace shapes of Offshore Main Coast are more similar in fig 9 to the Scotian Shelf and Georges Bank population.

Despite the relatively even distribution of ESD across the Gulf of Maine the observations are of relatively low frequency and it is a challenge to test whether these data represent a Poisson distribution of rare observations or if the observations are significantly clustered and reproducible. I will examine if the data are sufficient for such calculations in a subsequent post.

Bibliography

Berejnov V. (2009). Rapid and Inexpensive Reconstruction of 3D Structures for Micro-Objects Using Common Optical Microscopy, xxx.lanl.gov

Kunkel Joseph G, Melissa Rosa, Brian Tarbox, Sabine Hild, Michael J. Jercinovic, Ali N. Bahadur (2018). Recognizing Incipient Epizootic Shell Disease Lesions in the Carapace of the American Lobster, Homarus americanus, H. Milne Edwards 1837. Bull Mar Sci 94(3):863-886.

Rohlf FJ, D Slice. (1990). Extensions of the Procrustes Method for the Optimal Superimposition of Landmarks. Syst. Zool., 39(1):40-59.

Semyonov D. (2018). Metashape. Agisoft LLC, St. Petersburg, Russia. https://www.agisoft.com/forum/index.php?topic=9793.0

Smolowitz RM, Chistoserdov AY, Hsu A (2005) A description of the pathology of epizootic shell disease in the American lobster, Homarus americanus H. Milne Edwards 1837. J Shellfish Res 24:749−756

(Abstract submitted for 2021 National Shellfish of America virtual meeting Spring 2021)
Joseph G. Kunkel¹, John K. Galbraith², Jakub Kircun², Carl Wilson³ and Kathleen Reardon^3
1 University of New England and UMass Amherst
² NOAA National Marine Fisheries
³ Maine Department of Marine Resources
Epizootic Shell Disease (ESD) have been increasing in Homarus americanus in the Gulf of Maine. To quantify this we surveyed ESD during the NOAA Northeast Groundfish Survey and establish protocol consistency with inshore surveys. Protocol was inserted into the NOAA FSCS sampling routines recording ESD in lobsters using NOAA and volunteer scientists during the two groundfish surveys per year. Each trawl station with lobsters was entirely evaluated or subsampled. Suspect ESD lobsters were barcoded recording metadata: location, sex, carapace length and total weight, noting any loss of legs and chela, whether egg bearing, tail notched, and barnacle presence. Carapace ESD was graded estimating coverage. The barcoded carapace of all ESD(+) were frozen for reevaluation on land and photogrammetry. This resulted in 3-D surface models of the carapace with descriptive metadata for each lobster and identity as true ESD or false(+). Data were found to be consistent with the general observations of prior ESD studies: (1) females were more abundant in the ESD(+) category and typically had a higher level of ESD. (2) Lower ESD(+) was seen in Fall surveys, higher ESD(+) in Spring. We conclude that the female ESD(+) class is substantially undercounted because the level 3 ESD should include deceased female lobsters.

The carapace of a normal lobster is sleek and slippery to the touch: Normal Adult Female. The slipperiness of the lobster to the touch is based on an unstirred layer of dissolved calcite, a crystaline form of CaCO₃. The dissolved carbonate takes a proton from water becoming bicarbonate and leaving a hydroxyl which can raise the pH of the unstirred layer to be pH 9 or 10.

CaCO₃ + H₂O = Ca²⁺ + HCO₃^-1 + OH^-1

This high pH layer inhibits the life functions of bacteria which find it hard to settle and form a biofilm on the lobster carapace. Motility of bacteria depend on the bacterial flagellum which is a rotary engine driven by protons. High pH turns off the source of energy for this Wankel engine. Many nutrients are co-transported into the bacterium with protons and high pH turns off bacterial nutrition.

The Spring 2018 NE Groundfish Survey, leg4, was spent collecting images of the American lobster that could be processed int 3D models of the lobster carapace. Specimens of ESD were sought for imaging as well as normal and other cuticle aberrations that can be confused with ESD and be counted as false positives of ESD. The images were processed with software Photoscan producing movies of the rotatin carapace allowing an observer to visually estimate the percent of the carapace surface involved in ESD lesions.

http://www.bio.umass.edu/biology/kunkel/pub/lobster/3D/PhotoScan/

Comments on the project and the carapace movies should be sent to Joe Kunkel at joe@bio.umass,edu .