Bug in the sign convention in the TKE OSBL scheme?

,

I am looking through the turbulent kinetic energy (TKE) ocean surface boundary layer (OSBL) scheme but I am having some confusions pertaining to the term associated with Kolmogorov dissipation. In the snippet of code from zdftke.F90 pasted below, the term with dissl has a + sign. Shouldn’t this be minus in order to ensure a reduction in TKE, en?

DO_3D_OVR( nn_hls-1, nn_hls-1, nn_hls-1, nn_hls-1, 2, jpkm1 )   !* Matrix and right hand side in en
   ...
   !
   !                                   ! right hand side in en
   en(ji,jj,jk) = en(ji,jj,jk) + rn_Dt * (  p_sh2(ji,jj,jk)                       &   ! shear
      &                                 - p_avt(ji,jj,jk) * rn2(ji,jj,jk)          &   ! stratification
      &                                 + zfact3 * dissl(ji,jj,jk) * en(ji,jj,jk)  &   ! dissipation
      &                                ) * wmask(ji,jj,jk)
END_3D

The sign of dissipation also seems to contradict Equation (9.1) in the NEMO reference manual.