Influences of climate and fertiliser application history on various measures of soil fertility and productivity in Wairarapa hill country : a thesis presented in partial fulfilment of the requirements for the degree of Master of Agricultural Science in Soil Science at Massey University, New Zealand

During the agricultural downturn of the mid 1980's, it became uneconomic for many Wairarapa farmers to apply fertiliser. Those farmers who could afford to apply maintenance fertiliser, often chose to apply diammonium phosphate (DAP). The increasing popularity and apparent effectiveness of fertilisers such as DAP, which has a nitrogen component, prompted further questions of whether traditional P and S fertiliser application (as single superphosphate), applied to stimulate pasture legume growth and biological nitrogen fixation, was the most cost effective fertiliser strategy. A review of literature revealed that soil fertility and pasture production data for different rainfall regimes in the Wairarapa was scarce. No published data was available on how effective single superphosphate (SSP) applications to Wairarapa hill country farms had been in increasing annual N fixation rates and plant­ available soil N, and hence increasing pasture production. It was concluded that it was necessary to conduct a series of pasture field trials on sites varying in soil type, fertiliser history and climate (annual rainfall) in order to provide data applicable to the Wairarapa region. Sites falling within three rainfall regimes were selected, those being Mauriceville (high rainfall), Gladstone (summer dry) and Whareama (low rainfall) sites. Using total soil phosphorus, sulphur and nitrogen analyses, a total of 14 sites were selected for study, with sites ranging in fertility status within each climate zone. The objective has been to characterise the soil types, soil nutrient status, climate and current pasture production of Wairarapa hill country farms, with a view to completing further studies examining in more detail the complex interactions of soil, climate and pasture. Soil chemical analyses revealed a wide range of soil fertility status across all sites. Soil total phosphorus (P) contents ranged from 430 µgP/g soil (site 4) to 1470 µgP/g (site 1), while total soil sulphur (S) showed less variation, ranging from 345 µgS/g (site 11) to 860 µgS/g (site 9). Soil total nitrogen (N) contents followed a similar pattern to that seen for total S, and ranged from 4280 µgN/g (site 11) to 7950 µgN/g (site 9). Fertiliser history had a large influence on the accumulation of P, S and N in these hill country soils, where higher rates of accumulation were associated with greater levels of fertiliser input. However, P accumulated at a far greater rate than S and Nin these soils, which is possibly the result of high S and N leaching losses. Measurements of plant-available nutrient levels followed a similar trend to that seen for soil total elemental analyses, where higher levels of nutrient were correlated with higher fertiliser inputs. Estimates of the efficiency of past fertiliser applications were made for these sites, using the results of various soil analyses. Traditional P and S fertiliser applications and pasture management on Wairarapa hill country appear to have been particularly inefficient in causing soil N to accumulate. The calculations used to derive these estimates have limitations, but do indicate that either product N and leaching losses are high or N fixation rates are low (or both) in these soils. July 1993 to March 1994 was an average growing season, and total herbage yields harvested from August 1993 to March 1994 ranged from 4 tDM/ha (site 4) to 15.5 tDM/ha (site 1). Soil fertility status (and hence historical SSP applications) was the main factor influencing total herbage yield, where high yields were recorded at high fertility sites, and the reverse for low fertility sites. Climate (soil moisture levels) also influenced total yield but to a lesser extent than soil fertility status. Pasture growth at Gladstone and Whareama sites stopped when soil volumetric water content in the top 7.5 cm fell below 0.2. At the wetter Mauriceville sites, soil moisture was not limiting until mid-February 1994. Legume growth was particularly sensitive to soil moisture stress. By converting pasture production to stock units, gross margin analyses were performed. The most profitable options in all three rainfall regimes were sites which had received frequent fertiliser applications. This suggests that historical fertiliser applications are economically effective, which is an important factor in sustainable agricultural enterprises. Herbage N and P uptake results supported this finding, and showed that pasture N uptake varied widely between high and low fertility sites. Pasture N uptake ranged from 70 kgN/ha at low fertility site 4, to 250 kgN/ha at high fertility site 1 for the period of early August 1993 to early January 1994. This implies that historical superphosphate applications have been effective in providing large increases in annual amounts of plant-available soil N at high fertility sites when compared to unfertilised sites, despite the fact that soil N accumulation was less than expected. Acetylene reduction activity (ARA) measured at each harvest showed that annual N fixation levels are limited by extended summer dry periods which stop legume growth. The wetter Mauriceville sites fixed more N on an annual basis than Gladstone and Whareama sites. ARA was linearly related to yield. Variations in the data indicated that species and other short-term soil condition changes have a large effect on the relationship between ARA and N fixation rates. Although soil N accumulation is slow in these pasture systems, annual pasture N uptake is dramatically increased where fertiliser inputs have been high. The results indicate that rapid cycling of soil/plant N is occurring, and that annual leaching and product losses of N may nearly equal N fixation rates. This was exemplified in a simple budget of the nitrogen cycle, taking account of N losses and gains in a low fertility and high fertility system. There was insufficient information to conclude why soil N is not accumulating in these grazing systems. Further research is required to fully explain this N cycle, including the relative quantities of N inputs and losses from the system