It's a delicate dance to avoid DO-issues once biomass and water-temps reach critical levels. Killing a large biomass in cooler water is equally as risky as killing a smaller biomass in warm water. The only way to REDUCE the risk of DO problems is to avoid both scenarios.
In my experiences, DO-crashes are "mostly" due to large-scale die-offs of filamentous algae populations; but I've also caused a fish-kill (both predicted and client-authorized) during an aggressive coontail treatment in mid-summer.
For your described situation, I suspect that treating a milfoil population that grows primarily in only a section of the pond - early season - is unlikely to induce a DO problem; but no guarantees.
Sub-surface injection can be accomplished with many home-made setups, and your 25-gal ATV-type sprayer is well suited for such purposes. Replacing the spray-gun with a perforated lead-filled pipe affixed to the end of an extension-hose should suffice.
The label-rate for 2,4-D is somewhat ambiguous and misleading. I've had good results (on milfoil) with far less product than the label prescribes; but it also depends upon several site and treatment-factors. Probably best to discuss this aspect of any potential treatments.
Staggered whole-bay treatments are preferable, and shouldn't pose a DO issue throughout the entire lake. The demise of milfoil; post-treatment, will be gradual enough that DO isn't likely to cause an issue - in most cases.
Lake dyes probably aren't a viable option in ponds that have perpetual water-transition - depending on the rate of transition and your pocketbook. Even then, a lake dye only "works" when there's adequate depth available for sunlight-filtration. Same thing applies to fertilization. And, for that practice to work, other water-quality factors must be suitable for the fertilizer to produce a adequate phytoplankton bloom.
I'll shoot you an email with my ph#. Maybe we can dialog one evening this week concerning site-factors and potential treatment options/methods.
