Design Process

Although crab is a benthic, and thus bottom feeding, species, the blue crab adult is an inefficient eater and tends to leave behind food remains that are too small to consume. For this reason, it was suggested early in the process that the project pair blue crab raceways with shrimp raceways, circulating water between the two such that the shrimp consume the remains of the food left by the crabs. This association was a stepping off point. Along with the requirement for a fallow raceway for water discharge/recharge, this pairing began to inform the size/scale of the raceway component of the project.
Since recirculating aquaculture can be performed almost anywhere and at almost any scale, the first site parti was to imagine the largest, most generic production facility—one that would ignore both the particulars of the site and the requirements imposed by the proposed location in Dorchester County, MD. This site scheme was largely an adaptation of the Thanet Earth compound in the UK (fig. 42) which houses over 1.3 million hydroponic plants in 220 acres of glass enclosed greenhouse.

Figure 42 - Thanet Earth compound. Author illustration. Photos courtesy Thanet Earth
The problems with this first scheme are numerous and readily apparent. A glass enclosed complex of this magnitude would have difficulty venting stale air and moderating temperature. What is more, the vast footprint of the glassed in area will not allow for easy rainwater collection outside the building enclosure (fig. 43).

Figure 43 - Site Scheme 1. Author Illustration.

The second site scheme made an attempt to break down the scale of the system, while keeping the pairing of crabs/shrimp intact (fig. 44). This scheme included a central stormwater retention/wastewater remediation feature, but the immense cut/fill required to maintain positive drainage to a central cistern quickly rendered this approach impractical. As a result of this second scheme, the third design focused on breaking down the scale of the greenhouse even further, and localizing rainwater retention/detention features.

Figure 44 – Site Scheme 2. Author illustration.

Site Scheme 3 revealed the promise of remediating fecal waters in localized wetlands, but, as before, its size was predicated upon the crab/shrimp pairing that conjoined two crab raceways with a shrimp raceway, thus tripling the necessary width of the enclosed space and necessitating a more complicated greenhouse solution (fig. 45).

Figure 45 – Site Scheme 3. Author Illustration.
After further research, it became clear that the financial value of the particulate food matter the crab cannot ingest was of small value compared to the costs of the other inputs in the system. Thus, in Site Scheme 4 the crab/shrimp association was eliminated (fig. 46). This scheme attempted to determine the smallest size/scope of a facility that would employ 30 fulltime pickers year-round. This fourth scheme in many ways epitomizes the traditional thinking behind agricultural farming. In this scheme the existing loblolly forest is clear-cut to allow for a single-harvest monoculture transplant into the surrounding ecosystem. As such, there is immense biodiversity at the farm’s periphery, but none at its core. The flora and fauna that once populated this site are exiled to the forest’s edge, and the farm itself becomes an island of homogeneity in a sea of biodiversity. After reflecting upon this scheme closely, it became clear that herein lay the real dynamic of the thesis: creating an artificial exosystem without seriously impairing the existence of those species displaced by the artificial transplantation.

Figure 46 - Site Scheme 4. Author illustration.


Site scheme 5 was a major leap forward (fig. 47). This scheme attempted to disaggregate the crab raceways such that the loblolly pine forest penetrated the site as a series of greenways that could allow animal movement along the periphery of smaller more amorphous clusters of raceways associated with local constructed wetlands for stormwater retention and wastewater remediation.

Figure 47 - Site Scheme 5. Author illustration.

Upon review, this strategy suffered in a few ways:

1) the greenways were not wide enough to accommodate real animal movement whether it is seasonal or migratory. According to empirical research, the width of a movement corridor is a matter of tens of meters. It’s therefore likely that 42 feet (1 program bay) would not be wide enough to allow for real animal movement.

2) The wetlands did not locate at the periphery of the raceways clusters such that they shared adjacency with the greenways. Animals drink from these wetlands (the salinity content in parts per thousands is not so great that foxes/deer/other fringe species cannot safely digest it), therefore it would be wise to pair these two features, at least where possible.

3) This scheme would require a clear cutting of the site (to berm site soils and replant the loblolly greenways at a raised elevation), thereby eliminating any of the 30+ year old forest canopy (a non-starter).

4) This scheme does not provide adequate workspace for the cleaning of the raceways after they have been emptied of their water and their crab contents. There would be a real need for open space, or “lawn” elements adjacent to each of the raceways that could be used to abandon plant and regrow the aquatic plant life which would act as habitat during the next growing season.

5) This scheme does not take into account the extremely long shadows thrown by the loblolly trees which grow upwards of 80-100 feet and sometime reach even 120+ feet. The glass enclosed raceways therefore cannot be located directly to the north of the greenway elements.

6) This scheme employed raceways of different sizes in an effort to feather the edge of the raceway clusters. This level of complexity is ultimately unnecessary because the same goal could be achieved using a single raceway prototype and modifying the grouping of these raceways to achieve the same goal.

Site Scheme 6 aimed to remedy the shortcomings of the fifth scheme, but ended up regressing in several regards. While it did make up for a few of Scheme 5’s deficiencies, it was problematic in terms of both the truck routing and the aesthetics of the experience (fig. 48). The relocation of the building from the northeast corner down to the midpoint of the site was not beneficial, because it only complicated the movement of trucks about the site, particularly in the northwest corner. The determined effort to render the raceway clusters more figural in their general outline and shape was equally misdirected. But it was upon viewing this solution which finally drove home that it was not the delineation between raceway (figure) and greenway (ground) that should be pursued, but rather the blurring of this distinction.
Figure 48 - Site Scheme 6. Author illustration.

As such, Scheme 7 would feature both an afigural grouping of raceways and a feathering or overlap of forest and raceway that would lend the final scheme its mosaic quality.

Site Scheme 7 made a determined effort to envision the greenway as figure, not the raceway as figure (fig. 49). As the greenway was thickened and the rigid hierarchy of east-west streets was broken, the forest congealed to the east and the west of the site, and the greenway began to balance the raceway clusters in size. The greenway never fully became figure in the ground of the raceway clusters, but the effort resolved many of the issues that had detracted from earlier schemes. Scheme 7 resolved the lingering truck routing issues by adding a secondary north-south roadway to the west of the honorific (visitor) axis. This gesture successfully closed the truck routing circuit in such a way that the trucking would be performed on the site in a clockwise fashion. Instead of breaking the honorific “street” every 42’ with a roadway, the service roads begin to aggregate and collect the east-west service roads such that these roadways merge and cross the honorific axis at only four locations. A similar strategy of aggregation was used at the east of the site to avoid penetrating/paving the forest where unnecessary.

Figure 49 - Site Scheme 7. Final product of iterative process. Author Illustration.

[34] Foreman, 155.