The definition of a reflexive relation over $A$ is:
$R$ is reflexive over $A$ iff $\forall a \in A :(a,a) \in R$
Why the '$\forall a \in A$'? Def. of transitive and symmetric relations don't have that:
$R$ is transitive iff $(a,b)\in R \wedge (b,c) \in R \implies (a,c) \in R$
$R$ is symmetric iff $(a,b) \in R \implies (b,a) \in R$
If the universal quantification is needed to define the reflexive relation, why isn't it needed for the other two?
Strictly speaking, the universal quantification should be there in all three cases. It's often just left out as a shorthand.
Edit: strictly speaking, all statements are implicitly universally quantified. This is true even if the statement is written as an implication e.g.
$$\text{ if } a \in A \text{ then } (a,a) \in R$$
is really just
$$\forall a \in A, (a,a) \in R$$
or if you like, it is equivalent to the redundant statement:
$$\forall a \in A, \text{ if } a \in A \text{ then } (a,a) \in R$$
Otherwise, what is the $a$ that we are talking about? It needs to be drawn from some universe. So much of maths is implicitly universally quantified that we forget it's going on in the background. But from a strict, logical standpoint, it's always there.