Dystroglycanopathies include the most severe forms of congenital muscular dystrophy (CMD) also associated with heterogeneous brain and ocular malformations. These disorders are caused by mutations in several genes that converge on the synthesis of a specialized glycan on alpha-dystroglycan (alpha-DG), the transmembrane component of the dystrophin-glycoprotein complex. Mouse knock-out (KO) models for dystroglycan itself and enzymes involved in the initial steps of glycosylation lead to early embryonic or perinatal lethality requiring conditional removal. Our long-term goal is to leverage the zebrafish as a model of muscle, eye, and brain disease to study genetic and phenotypic heterogeneity in dystroglycanopathies. Here, we introduce two zebrafish KO lines for enzymes that initiate alpha-DG glycosylation, protein O-mannosyltransferase 1 (pomt1) and protein O-linked N-acetylglucosaminyltransferase 2 (pomgnt2). Previous work showed extreme phenotypic variability in dystroglycanopathy zebrafish models, with some mutants surviving to adulthood with only retinal deficits. We reconcile these findings by showing the critical role of maternal mRNA in masking developmental phenotypes by retaining alpha-DG glycosylation. Both pomt1 and pomgnt2 mRNAs are provided by the mother in the oocyte, and pomt1 and pomgnt2 KO larvae from heterozygous parents (termed KOHet) retain glycosylated alpha-DG during the first weeks of life. Thus, KOHet lines exhibit prolonged survival with juvenile- to adult-onset muscle disease and locomotor deficits. In contrast, when maternal pomt1 or pomgnt2 is depleted by breeding KOHet females, the resulting KOs (termed KOKOs) exhibit profound, larval-onset muscle, brain, and eye phenotypes and reduced survival. RNA-seq studies on larvae showed minimal differential expression between KOHets and controls, while KOKOs have downregulation of genes involved in muscle formation and contraction. We have further outlined fundamental differences in how these glycosylation events affect the neuromuscular junction. These new models can now be used to study pathogenesis in multiple tissues and novel therapeutic interventions.