Harvest cycles as a design input, not an afterthought

Most orchards are designed for planting and discover their harvest problem later. The trees go in by what grows well together and what the nursery had in stock, and then one November the whole place ripens at once — a glut no household can eat, no local market can absorb at a fair price, and no smallholder has the labour to pick. Harvest cycles are not an afterthought to be managed at the end; they are a design input to be planned at the beginning. Choose your species by when they pay you, not only by what they yield, and you design the glut out before you ever dig a hole.

The glut is a design failure, not a season

Toby Hemenway’s framing is that a well-designed system delivers a steady flow of yield, not a single flood. The single-November-glut orchard fails on three fronts at once: it overwhelms household consumption, it floods the local market exactly when everyone else’s identical trees are also ripe (so prices crater), and it concentrates all the labour demand into the few weeks when family labour is most stretched. Every one of those failures is fixable at the design stage by spreading the harvest across the calendar — and none of them is fixable after the trees are mature and synchronised.

Stacking the calendar with Pakistani species

The Pakistani agroforestry palette can fill almost the whole year if chosen for spread. Citrus carries the cool months (kinnow and oranges through winter). Loquat and early stone fruit open the spring. Mango defines the summer (and within mango, staggering Sindhri, then Chaunsa, then late varieties spreads even the mango window across weeks). Guava bears across two seasons. Dates, jujube (ber), and pomegranate carry late summer into autumn. Olive and walnut add autumn harvests in the cooler zones. Lay these on a twelve-month grid and the design question becomes obvious: which months are empty, and which species fill them? Plant to flatten the curve, not to maximise any single peak.

Designing for labour, not just for fruit

The harvest calendar must be read against the labour calendar. Pakistani smallholder labour peaks in the cooler October-to-March window and drops in the brutal June-to-August heat — and mango, the headline crop, ripens squarely in that low-labour window. A planting that piles still more harvest into June-August compounds the labour crunch; one that pushes yield toward the shoulder seasons works with the household’s actual capacity. Sequencing species so that harvests fall when hands are available is as much a part of the design as sun and water.

Post-harvest is part of the cycle

Spreading the harvest only helps if what you pick survives to be sold. Pakistani smallholder fruit suffers heavy post-harvest losses — bruising, heat, and the absence of any cold chain turn a good crop into a discounted one within days. Two design choices blunt this: stackable harvest crates that prevent the bruising of sack-and-pile handling, and solar drying that converts a surplus that would rot into a storable, higher-value product (dried mango, dried jujube, dried chilli) sold months after the fresh glut has passed. Designed in from the start, drying capacity also flattens the income curve — turning the unavoidable peak into product you sell across the lean months. Harvest-cycle design, done fully, reaches past the picking into how the yield is held, moved, and sold.