Ship biofouling is a major vector for the introduction and spread of harmful marine species globally. Comprehensive underwater sampling and video recording of ship hulls was conducted to assess biofouling extent (percent cover, total abundance and species richness) on a subset of ships arriving to Canadian waters. The dataset includes underwater biofouling assessments from 53 commercial ships arriving at Halifax, Nova Scotia (20 international ships), Vancouver, British Columbia (20 international ships) during 2007–2009, and Churchill, Manitoba (11 international and 2 domestic ships) in 2010–2011. Potential explanatory variables in the dataset include ship size, typical sailing speed, port residence time, age of antifouling coating system and travel history (number of biogeographic realms visisted, and average, minimum and maximum port latititude).

Underwater surveys were conducted by professional divers while ships were stationary in port. Divers surveyed both sides of the hull of each ship from bow to stern and bottom to waterline, although access to the mid-ship section was normally limited by the size of the gap between the hull and the berth bed and wall. The rudder sides, bottom, leading and trailing edges, propeller nose and blades, rope guard, stern tube, sea-chests, bow-thruster tunnel and grating, bulbous bow, stem and main hull were inspected for every ship. Physical samples were collected at each location where growth was observed, whereas a value of zero organisms was recorded in locations where no growth was found. Physical sampling was not random, but aimed to include the highest possible number of organisms and species. Average abundances m^2 and for the whole ship were estimated using percent cover information from random video-transects of the entire hull. Sampling was conducted at each location with 1–3 replicate 20 · 20 cm magnetic quadrats attached to the hull of each vessel, or equivalent surface area at uneven locations where the use of the quadrat was not practical. Barnacle samples were scraped into re-sealable plastic bags and a suction device mounted on a scraping blade was used to collect soft growth. Two-L water samples were also collected from the dock at mid-hull depth as controls. The water volume of each hull fouling sample was measured at the dock using a graduated cylinder, and organisms in the equivalent volume of control water sample were subtracted from abundance counts (see below). Species present in control water samples were subtracted from species richness estimations. For a subset of ships, hull samples were examined by eye at the dock, prior to fixation, to determine if organisms were dead or alive when collected. However, this analysis was limited to barnacles, amphipods, chironomids, mites, large bivalves and gastropods as remaining groups were not large enough to be checked reliably. Specimens that appeared alive (e.g. movement, valve/plate closure) were preserved separately. All samples were sieved through a 40 um mesh and preserved in 95% ethanol at the dock. Information including the list of ports-of-call since last dry-dock or 1 year previous to sampling, antifouling protection and the sheet of particulars were collected from the crews of each vessel. All samples were processed in the lab to estimate abundances and conduct taxonomic identifications. Macroalgae were not present in samples. Instead, algal cover was restricted to relatively thin films lacking reproductive structures, thus only organisms from selected samples were identified at a coarse taxonomic level. For this reason, efforts were focused on invertebrate animals.

The in-water sampling design did not allow divers to open and sample inside sea-chests and other protected locations, although sampling was conducted from the grating covering the sea-chests. Therefore, hull fouling was probably underestimated in such locations.