Invariant and holomorphic distributions on para-Kenmotsu manifolds (Q895811)

The author deals with two questions on para-Kenmotsu manifolds [\textit{K. Kenmotsu}, Tohoku Math. J., II. Ser. 24, 93--103 (1972; Zbl 0245.53040)]. One is the characterization of holomorphic vector fields as the kernel of a \(\overline{\partial}\)-operator. The second one is the description of the Walczak formula [\textit{P.G. Walczack}, Colloq. Math. 58, No. 2, 243--252 (1990; Zbl 0766.53024)] for certain pairs of orthogonal distribution on para-Kenmotsu manifolds. The paper is recommended to those researchers interested in (almost) para-contact geometry and foliations from the geometric point of view. The paper is technical but well written, although it should be noticed that it lacks examples and motivation. The first and second sections introduce basic definitions and properties. In the third section the author defines the notion of holomorphic vector field (and distribution) and shows that the set of holomorphic vector fields is a Lie subalgebra. One of the relevant results of the paper shows that the set of holomorphic vector fields on an almost paracontact manifold \(M\) with structure \((\varphi,\xi,\eta,g)\) agrees with the kernel of the operator \(\overline{\partial}:\Gamma(TM)\to\mathrm{End}(TM)\): \[ \overline{\partial}X(Y)=\varphi\biggl(\nabla_Y X-\varphi(\nabla_{\varphi Y}X)+\varphi((\nabla_X\varphi)Y)\biggr). \] As it is noticed in the paper, this must be understood as an analogue of what happens in the complex and contact cases. Additionally, the author gives a characterization of some holomorphic vector fields on the cone over a para-Kenmotsu manifold. In the fourth section the integrability of a \(\varphi\)-invariant distribution \(\mathcal H\) is studied in terms of its second fundamental form, called \(B^{\mathcal{H}}\), and its integrability form, called \(I^\mathcal{H}\), both relative to the orthogonal distribution \(\mathcal V\). The main result in this section shows that \(\mathcal H\) is integrable if \(\xi\in\Gamma(\mathcal H)\) and \(\varphi\) completely commutes with the second fundamental form \(B^{\mathcal H}\). The last section is the most technical one and deals with the Walczak formula for orthogonal distributions. After several pages of computations the author gets a beautiful formula for the volume of \(M\) in terms of a pair of orthogonal distributions \((\mathcal V,\mathcal H)\) so that \(\mathcal V\) is \(\varphi\)-invariant, holomorphic and both distributions are integrable and umbilical: \[ q(q-1)\mathrm{vol}(M)=\int_M \biggl(K(\mathcal{H},\mathcal{V}) - (p-1)\|B^{\mathcal V}\|^2\biggr)d\mu_g, \] where \(p=\dim\mathcal V\) and \(q=\dim\mathcal V\), \(p+q=\dim M=2n +1\), \(\mu_g\) is the Lebesgue measure associated to the pseudo-Riemannian metric \(g\), \(\|\cdot\|\) is the Hilbert-Schmidt pseudo-norm and \(K(\mathcal H,\mathcal V)\) denotes the mixed scalar curvature. Notice that \(\mathcal H\) is umbilical iff \(g(B^{\mathcal H}(X,Y),Z)=g(\omega(Z)X,Y)\), for \(X,Y\in\mathcal H\) and \(Z\in\mathcal V\). Whether these hypotheses are reasonable is not discussed in the paper.