[en] The trans-Neptunian objects (TNOs) preserve evidence of planet building processes in their orbital and size distributions. While all populations show steep size distributions for large objects, a relative deficit of Neptunian trojans and scattering objects with diameters of D < 100 km has been detected. We investigated this deficit with a 32 square degree survey, in which we detected 77 TNOs that are brighter than a limiting r-band magnitude of 24.6. Our plutino sample (18 objects in 3:2 mean-motion resonance with Neptune) shows a deficit of D < 100 km objects, rejecting a single power-law size distribution at >99% confidence. Combining our survey with the Canada–France Ecliptic Plane Survey, we perform a detailed analysis of the allowable parameters for the plutino size distribution, including knees and divots. We surmise the existence of 9000 ± 3000 plutinos with an absolute magnitude of H _r ≤ 8.66 and ${37000}_{-<!--\; ???\; -->10000}^{+12000}$ with H _r ≤ 10.0 (95% confidence). Our survey also discovered one temporary Uranian trojan, one temporary Neptunian trojan, and one stable Neptunian trojan, for which we estimate populations of ${110}_{-<!--\; ???\; -->100}^{+500}$, ${210}_{-<!--\; ???\; -->200}^{+900}$, and ${150}_{-<!--\; ???\; -->140}^{+600}$ with H _r ≤ 10.0, respectively. All three populations are thus less numerous than the main belt asteroids (592 asteroids with H _r ≤ 10.0). With such population sizes, the temporary Neptunian trojans cannot be previously stable trojans diffusing out of the resonance now; they must be recently captured Centaurs or scattering objects. As the bias against the detection of objects grows with larger semimajor axes, our discovery of three 3:1 resonators and one 4:1 resonator adds to the growing evidence that the high-order resonances are far more populated than is typically predicted.