Measuring the impact of invasive N-fixing Acacias on nitrogen dynamics in fynbos riparian zones of the Western Cape*

Riparian zones in the fynbos biome are especially vulnerable to invasion by non-indigenous woody vegetation Fast growing Acacia mearnsii is able to suppress fynbos through its ability to outcompete fynbos for sunlight, water and nutrients. The current study also suggests that nitrogen-fixing A. mearnsii is able to affect fynbos ecosystem processes by changing the nitrogen cycle. SAEON MSc student Marno Fourie

By Marno Fourie, SAEON Student

In the fynbos biome, riparian zones are especially vulnerable to invasion by non-indigenous woody vegetation that can cause degradation to these habitats by suppressing or replacing indigenous vegetation.

In addition to changes in species composition, processes in the riparian zone such as nutrient recycling may also change. There is an assumption that riparian zones will spontaneously repair, both structurally and functionally, once invasive vegetation is removed.

We studied the impact of Acacia mearnsii, a dominating, invasive, nitrogen-fixing tree species on aspects of the nitrogen cycle. Seven perennial river systems in the south-western Cape (Figure 1) were studied to determine whether soil nitrogen-cycling processes (e.g. ammonification, nitrification and denitrification; Figure 2) are restored to the pre-invasion state following clearing.

The study expected to find higher nitrogen mineralisation rates and higher levels of available nitrogen, especially in the form of nitrates, in the riparian zones where Acacia mearnsii was the dominant vegetation type compared to reference sites. Areas that were cleared of A. mearnsii were expected to recover in terms of decreased nitrogen mineralisation activity, higher levels of ammonium and the reduction of nitrates through the process of denitrification. However, nitrogen dynamics were not quite what we expected.

Nitrogen stocks in the form of total nitrogen, organic nitrogen and inorganic nitrogen increased in riparian zones invaded by A. mearnsii to levels that exceeded those of reference (uninvaded) sites. Clearing of A. mearnsii resulted in higher nitrogen stocks than invaded areas, and higher than at the reference sites, even though clearing took place seven years prior to commencement of the study.

Rates of nitrogen mineralisation declined and reached negative levels as a result of the invasion by A. mearnsii. This suggests that nitrogen is bound up in organic forms. However, following clearing of A. mearnsii, nitrogen mineralisation increased, although not to pre-invasion levels. Invaded sites contribute towards higher levels of inorganic nitrogen compared to reference sites and clearing of A. mearnsii causes inorganic nitrogen to decrease to levels closer to those of reference sites (Table 1).

Actual immobilisation levels of inorganic nitrogen for invaded sites were higher than reference sites. Clearing A. mearnsii leads to lower immobilisation levels (Table 1) and, as a consequence, organic N at invaded sites is higher. The same trend was found for potential immobilised inorganic nitrogen (Table 1).

The levels for potential immobilised nitrogen was lower than the actual immobilised nitrogen, which means that there is less nitrogen bound in the organic form under optimal laboratory conditions than under field conditions. The lower potential immobilised nitrogen levels indicate that underlying soil processes are responsible for reducing the availability of inorganic nitrogen.

Total available nitrogen in riparian zones is driven by nitrate although ammonium becomes more dominant over time, especially in winter when the ratio of ammonium to nitrate is higher in the dry banks (Figure 3).

Figure 1: Research sites represented by symbols with different colours: green ("natural" sites), red (densely invaded sites); and blue (cleared sites), UE: Upper Eerste River, LE: Lower Eerste River, UD: Upper Dwars River, LD: Lower Dwars River, EL: Elands River, WIT: Wit River, UM: Upper Molenaars River, LM: Lower Molenaars River, LJ: Lower Jakkals River; SL: Sir Lowry’s River, UJ: Upper Jakkals River

The position in the landscape (wet and dry banks in the riparian zone and upper terrestrial sites) did not have an effect on available nitrate when riparian zones were compared to the terrestrial areas beyond the riparian zone. Ammonium levels were different and increased across three seasons (from spring 2011 to autumn 2012 and winter 2012), especially in the dry banks of the invaded and cleared sites (Figure 4).

Figure 2: Conceptual soil nitrogen cycling model

The results show that the invasion by non-indigenous woody vegetation has an impact on riparian zones by increasing nitrogen input to the system, which was expected. It is clear from the negative rates of nitrogen mineralisation in the invaded sites that A. mearnsii reduces the biological cycling of nitrogen in the system and binds inorganic nitrogen in the form of organic nitrogen, which was unexpected. A. mearnsii invasion suppresses fynbos with their ability to fix nitrogen, grow fast and utilise resources such as sunlight, water and nutrients before the slow growing fynbos can be established.

Table 1: Summary of the nitrogen stocks and associated processes involved in the cycling of nitrogen, the reduction thereof and the maximum denitrification potential control (DEA), amended with carbon (DEAC), amended with inorganic nitrogen (DEAN) and amended with both carbon and nitrogen (DEA C+N) for the upper 25cm of the soil during spring in the Western Cape.

Areas that were cleared of invasive vegetation showed a recovery in nitrogen recycling processes in terms of increased nitrogen mineralisation activity, which was expected. This may eventually allow the system to recover to pre-invasion nitrogen levels, but before nitrogen reaches pre-invasion levels there is a risk that weedy, nitrogen-loving species (such as cosmopolitan weeds and grasses) colonise and outcompete native fynbos vegetation for resources. In addition, they may increase the frequency of fire, which is not suitable to fynbos life cycles.

Figure 3: Ammonium to nitrate ratio for A: invasion status (natural, invaded and cleared) and B: landscape position (wet bank, dry bank and terrestrial) across seasons. Data points are offset relative to time of sampling to aid in distinguishing the values associated with invasion status and position. The mean values of all data collected irrespective of invasion status or landscape position are depicted in graphs C and D. Mean values are indicated by different symbols and whiskers represent one standard error. Letters denote significant differences determined by repeated measures ANOVA followed by post hoc Fisher LSD tests (p < 0.05)

Figure 4: Ammonium for A: invasion status (natural, invaded and cleared) and B: landscape position (wet bank, dry bank and terrestrial) across seasons. Data points are offset relative to time of sampling to aid in distinguishing the values associated with invasion status and position. The mean values of all data collected irrespective of invasion status or landscape position are depicted in graphs C and D. Mean values are indicated by different symbols and whiskers represent one standard error. Letters denote significant differences determined by repeated measures ANOVA followed by post hoc Fisher LSD tests (p < 0.05)

Conclusions

Nitrogen stocks were enhanced following clearing of Acacia spp. especially in dry banks, where invasion was most prominent. Nitrogen mineralisation activity shows recovery towards uninvaded conditions following invasive alien vegetation clearing after seven years.

Invasion of Acacia spp. may lead to a depletion of plant-available nitrogen due to negative rates of nitrogen mineralisation despite higher soil pools of N. Invasive alien vegetation stands increase actual immobilised nitrogen in the form of organic nitrogen that is released to the atmosphere as nitrous oxide gas.

The lower potential immobilised nitrogen levels indicate that underlying soil processes are responsible for measured levels of actual immobilised nitrogen and that denitrification as a pathway to reduce nitrates in fynbos riparian zones are underestimated.

* MSc Conservation Ecology Thesis. Supervisors: Dr S.M. Jacobs and Dr A. Rozanov, Stellenbosch University