There are three non-burning and non-mechanical techniques most commonly employed in harvest weed seed control (HWSC) systems: chaff dumping, chaff lining and chaff tram-lining. The highest adoption of HWSC is in the GRDC western region with an estimated 67% of all farmers undertaking at least one HWSC strategy in 2014.

Chaff dumping is the collection of the chaff fraction using a cart towed behind the harvester. The chaff in the cart is then dumped, usually in piles in the paddock. The chaff is then either burnt, grazed or left to decompose.

For chaff lining, the chaff and weed seeds are confined to a row directly behind the harvester using a narrow chute. The chaff and weed seeds are then left to decompose over time. To promote decomposition, the chaff lines need to be placed in the same location year after year by running the harvester on a controlled traffic system (CTF).

Chaff tramlining is a similar concept to chaff lining, but the chaff fraction is diverted from the chaff deck onto permanent wheel tracks in a CTF system. Wheel traffic creates a hostile environment that inhibits weed seed germination.

There has been a recent system change with more growers opting for leaving chaff in-situ to rot, whether in dumps or in lines, rather than burning. This investigation aimed to better understand invertebrate species and mice associated with these HWSC systems, which to date are poorly understood, in each of the five port zones: Albany, Esperance, Geraldton, Kwinana’s East and West.

Both invertebrate pests and beneficials are found in association with chaff in paddocks. However, the species composition depends on the location of the paddock. Paddocks located in the Albany port zone are more likely to have higher densities of pests associated with chaff, whereas paddocks located in the Kwinana East port zone had the least pest densities and highest numbers of seed-harvesting ants associated with chaff.

The most abundant beneficials were weed seed harvesting ants which were found in close association with chaff. As these ants predate on weed seeds and live in holes in the ground, it is unlikely they are using chaff as a refuge, it is more probable that they are foraging in the chaff. This trend presents an opportunity to investigate in greater detail whether seed-harvesting ants provide an economic benefit to growers by consuming weed seeds from HWSC systems left to rot-in-situ.

Pitfall trapping was able to capture a higher diversity of invertebrate species than direct sampling of chaff. As pitfall trapping did not rely on samplers locating an invertebrate, it is more likely to be a more accurate representation of species presence and diversity in a paddock. However, this technique does rely on invertebrates moving and falling into a pitfall trap.

Leaving chaff to rot-in-situ does not affect abundances of pests such as desiantha weevil. However, other pests do use chaff as refuges. For instance, pest beetles, Rutherglen bugs, European earwigs were found in association with chaff. It is unlikely these species are feeding on chaff, but rather are using chaff as a refuge. This survey was not able to determine if long term retention of chaff will increase abundances of these pests.

The methodology for surveying for invertebrates associated with chaff was tested in 20 paddocks in 2019, before wide scale surveys of HWSC systems commenced. Two methods for invertebrate assessments (pitfall traps and direct surveys) were undertaken in the Geraldton, Esperance, Kwinana West, Kwinana East, and Albany Port zones.

A total of 87 paddocks were surveyed during 2019-2020, with of which 79 were sampled in 2020. An effort was made to identify a similar number of HWSC systems per port zone for this study. However, some HWSC systems are under-represented in certain zones. For instance, chaff tram lining is more common in the Esperance port zone than in the Kwinana port zones.

Two times of sampling occurred: prior to planting and post planting when crops were at the seedling stage

All invertebrates at each sample time and from each survey method were collected and then identified to species. Each species was divided into functional groups of beneficials (i.e. species that predate on pests or harvest weed seeds) or pests (i.e. species that are known to feed on germinating crops).

Pitfall trapping

Pitfall traps were placed at least 50 metres from any fence line or vegetation. These consisted of 250 mL containers, dug into the ground so the top lip was level with the surface. Pitfall traps were placed in two rows and kept open for 7 days. Each row consisted of 10 pitfall traps placed at least 10 metres apart.

One row was placed adjacent to chaff (near), at a distance of 5 cm from chaff; the second row (far) was placed parallel to the first at a distance of 20 metres away. In paddocks with chaff lines or tram-lines, the chaff lines are spaced at least 6 m apart. The ‘far from chaff’ pitfall trap rows were located between chaff lines and were at least 3 m from a chaff line (Figure 2, note chaff lines without a row of pitfall traps adjacent are not depicted).

In paddocks with chaff dumps, the second row was located 20 m from any chaff. Due to paddock variation, most chaff dumps were less than 100 metres in length. In this case, pitfall traps were placed 10 metres apart 5 cm from the chaff (close) and at least 20 m away from chaff (far) on the same side of multiple chaff dump.

Direct surveys of pests associated with chaff

On the eastern side of each pitfall trap, a 0.1m² quadrat was placed.

All chaff and/or stubble was removed from the quadrats and all invertebrates collected and assessed.

If crops were at the seedling stage, they were assessed for any invertebrate damage adjacent to each pitfall trap and a count of plants along a 1 metre by 2 crop rows was recorded.

A visual estimate of the percentage of plants that have chewing or sucking damage was undertaken:

1. Chewing damage is characterised with portions of the leaf missing. This can indicate if pests are present and causing damage. Chewing damage was scored on a scale of 0 to 100; with 0 being no chewing damage, 50 being half of the plants with chewing on leaves and 100 if all plants had chewed leaves.
2. Sucking damage is characterised by silvering or bleaching of leaves. Sucking damage was scored on a scale of 0 to 100, with 0 being no damage, 50 being half of the plants with some sucking damage on leaves and 100 if all plants had sucking damage.

Statistical analysis

Results for all years are shown below. Paired t-tests were used to compare near and far assessments for each of the three chaff systems. Unbalanced ANOVA’s were used to compare the three chaff systems.