The DESI DR2 preference for a negative effective neutrino mass under ΛCDM is known to be related to its preference for dynamical dark energy. We make that relation explicit and quantitative. At fixed acoustic scale, the neutrino and CPL background responses admit a unique first-order null direction, (δw0,δwa) ≃ (+0.7,−2.7)eV−1, of slope −4.1. We derive this slope in closed form from the Friedmann equation (−4.10) and reproduce it with two Boltzmann codes (−4.12 with CAMB, −4.11 with class). It is stable at the 3% level under curvature, Neff, mass splittings and non-CPL w(z), and it nulls 95% of the neutrino signal in the DESI χ2 for |δΣmν| ≲ 0.15eV. A second, near-exact background degeneracy involves early dark energy: its response is anti-parallel to the neutrino response at cosθ = −0.999, so an admixture δfede ≃ 5 × 10−4 per 0.1eV—far below current CMB sensitivity—can erase or manufacture an effective-mass signal of either sign. Growth breaks both degeneracies: 70–85% of the neutrino CMB-lensing signature survives dark-energy marginalization, and a Boltzmann-code calculation shows that clustering dark energy weakens, rather than strengthens, the mimicry. Constraints use official likelihoods (DESI DR2, Planck plik lite and CamSpec PR4, Pantheon+): freeing (w0,wa) moves the effective mass from −0.096±0.056 to +0.003±0.076eV, while Pantheon+ pins w0 and leaves the partial residual −0.060±0.084eV that the geometry predicts. Nested-sampling forecasts show that amplitude-type observables saturate at σ(Σmeff ν ) ≃ 0.08eV; the degeneracy-breaking information sits in the high-redshift expansion rate and in scale-dependent growth