This study, funded by Natural Heritage Trust Funding grant “Water Quality in Gwydir Valley Watercourses (Moree)”, investigated use by waterbirds of on-farm wetlands (mainly irrigation storages, but also other artificial wetlands) on irrigated cotton farms in the Lower Gwydir Valley. 

On-farm storages now cover c.120 km2, about 1.13% of the landscape, and constitute 45% of the total area of natural and artificial wetlands in the Lower Gwydir Valley.

Between September 1999 and July 2001, 23 surveys were undertaken of 19 on-farm wetlands on 9 cotton farms. During the surveys, we recorded numbers of birds present on each wetland and any signs of breeding. Wetland characteristics that were recorded included the species-richness and seedling production of their seed bank.

In all 23 surveys, a grand total was counted of 42,495 birds, of over 45 species. No species was detected that has not been recorded on the natural wetlands of the Lower Gwydir Valley, and several of the rarer species known from those natural wetlands (e.g. brolga, black-necked stork, comb-crested jacana) were not seen on the on-farm wetlands. Four of the 8 waterbird species listed on Schedule 2 of the NSW Threatened Species Act (magpie goose, blue-billed duck, one freckled duck and one Australasian bittern) were detected on on-farm wetlands, but only magpie goose regularly, in low numbers and breeding.

The recorded waterbird community on the on-farm wetlands was dominated by ducks, geese and swans (anatids), which constituted 59% of the birds recorded. Those were followed by: pelican, darter and cormorants (pelecaniforms: 18%); herons, egrets, spoonbills and ibises (ciconiiforms: 12%); and coot, moorhen and swamphen (rallids: 6.5%). The four most abundantly counted individual species were all ducks.The wetlands supported no or only few waders, snipe and cryptic crake and rail species. 

Total numbers of counted waterbirds varied from count to count and year to year, without obvious seasonal regularity. Numbers of some waterbird species on on-farm wetlands fluctuated rapidly and profoundly; others were more stable in numbers. Variations in numbers suggest that the waterbirds were highly mobile, using on-farm wetlands as part of the dispersed system of wetlands in the Lower Gwydir Valley. 

We found consistent and significant differences between the sampled on-farm wetlands (even within a farm) in the numbers, densities and composition of the waterbird communities they carried. The 5 most bird-rich on-farm wetlands carried 10-30 times as many birds as the 5 most bird-poor wetlands. 

On-farm wetlands differed in the predictability with which species-groups of waterbirds could be found on them (called their fidelity). Nine of the 19 wetlands had a predictability of 1-in-4 or better of finding any species-group, whereas 5 of the wetlands had average fidelities of less than 1-in-8. 

Average waterbird species-group fidelity increased with waterbird density on on-farm wetlands. However, individual waterbird species differed strongly in their fidelity scores. Some, such as whistle-ducks, were very numerous but occurred only irregularly; thus they had low fidelities. Others, such as black duck, were persistently present and showed high fidelity.

Our data showed that waterbirds in general, and anatids and rallids in particular, were significantly more numerous and predictably present on those on-farm wetlands:

• featuring trees in the water, beds of aquatic vegetation, and shallow areas that formed mud islands as water level fell; and
• whose sampled seed banks were most productive of seedlings and most species-rich.

Cormorants, darters and pelicans occurred predictably only on on-farm wetlands with numerous trees, usually dead ones, in the water.

Very few waterbirds bred on the on-farm wetlands. Some colonial, tree-platform-nesting waterbirds (such as darter, cormorants, spoonbills, and a few herons and ibises) nested on some wetlands with standing trees in the water, in each year of the study. However, numbers of nests per wetland were always low (<60). Some nesting terminated when water level was drawn down. Even fewer waterbirds nesting on the ground or on aquatic vegetation or in tree hollows (e.g. ducks) produced detected clutches of young.

Extrapolating from our survey results to the whole system of on-farm wetlands in the Lower Gwydir Valley suggests that, although the on-farm wetlands constitute 45% of the Lower Gwydir Valley’s mapped wetland area (excluding those wetland areas flooding naturally but briefly), they support on average 24,000 waterbird individuals, probably only 1-5% of the Valley’s waterbird community and less than 0.5% of waterbird nesting numbers.

However, if characteristics of at least some on-farm wetlands could be modified, without unacceptably reducing their usefulness for irrigation, to produce wetlands that are more diverse, more shallowly sloping, and with more aquatic vegetation and trees, then the widely dispersed system of on-farm wetlands could contribute substantially more than they do now to the conservation of a diverse waterbird community in the Lower Gwydir Valley.