New Zealand Science Teacher

To graze, or not to graze?

10/03/2015

TIM LOGAN writes about his research into the protection of native plants, for which he won the 2014 Future Scientist Prize. 

Tim Logan is a Year 13 Darfield High School student who has spent the past two years investigating the effect of stock grazing on indigenous grassland species located on the Waimakariri Plains.

For this work, he received the Prime Minister’s Future Scientist Prize in late 2014. Read more about his award on New Zealand Science Teacher here.

Tim shares his research here.

Abstract

The objective of this investigation was to determine whether grazing by stock enhances survival of native prostrate plant species, and to investigate how soil depth influences the vegetation. A field survey was carried out between January and March 2014; investigating 20 randomly selected 1x1m quadrats in two predominantly exotic grassed sites located within the mid-Waimakariri floodplain (central Canterbury). Results from Boxplots and Multidimensional Scaling (MDS) Analysis of quadrats shows that the biotic variation between the grazed and ungrazed site is substantial, with grazed quadrats showing a higher diversity and dominance of native species.

Species Accumulation Curves indicate that exotic vascular species tend to be widespread and relatively low in diversity; whereas native vascular species tend to be low in dominance and high in diversity. Stratified sampling across four landforms (terrace tops, channel floors, north facing scarps, and south facing scarps), and soil depth results conclude that shallower soils tend to correspond with an increased native percent cover and a decreased exotic per cent cover.

Tim's photo of the un-grazed site: The Kōwhai Sanctuary.

Introduction

The mid-Waimakariri floodplain historically consisted of open tussock grassland with scattered low trees and shrubs (Armstrong, 1879) on shallow, drought prone soils, subjected to periodic flooding. Over the last 150 years of European settlement, burning, stop-banking, and the introduction of adventive exotic plants has modified indigenous plant communities through competition and displacement of native species by competitively superior exotic plants (Meurk, 2008). Traditional dryland grazing practices in New Zealand consisted of extensive pastoralism with a low to moderate stocking rate (O’Connor, 1982), which was compatible with the survival of many non-palatable native species.

However, recent shifts towards more intensive agricultural practices involving ploughing, irrigating, fertilising and high stocking rates are essentially “squeezing...herbaceous species between intensification on one side of the fence and dense, uncropped exotic grasses on the other” (Meurk & Greenep, 2003). Today, the remaining semi-natural grasslands are dominated by exotic sweet vernal (Anthoxanthum odoratum) and brown top (Agrostis capillaris) grasses and exotic shrubs and trees – gorse (Ulexeuropaeus) and Monterey pine (Pinus radiata).

(Left: Figure 1)

Remnant native vegetation e.g. kōwhai (Sophora microphylla), matagouri (Discaria toumatou), and silver tussock (Poa cita) are present only in low density. However, beneath the exotic grasses, native prostrate plants and nonvascular plants are present e.g. creeping pōhuehue (Muehlenbeckia axillaris), dwarf broom (Carmichaelia corrugata) (Figure 1) and woolly moss (Racomitrium pruinosum) (Meurk, 2008). Small plants like these make up a large portion of our total biodiversity (Meurk & Greenep, 2003) but the dependence of these species upon low stature communities makes them incredibly susceptible to weed invasions and agricultural intensification, posing difficult questions for their management.

Environmental stress and disturbance is required to facilitate the regeneration of native plants as it results in the low stature vegetation necessary to give native herbaceous species a chance (Meurk & Greenep, 2003). While the effects of grazing by stock on native grassland species has been investigated by Meurk et al. (2002) in areas like the Mackenzie Basin, no such research has been conducted on the Canterbury Plains.

The idea to investigate the effect of grazing was born a couple of years ago while I was on a field trip to the grasslands. I was accompanied by Landcare Research ecologist, Dr Colin Meurk, who showed me some of these incredible species and told me of their decline and the challenges they will face in the future. These plants are often overlooked on account of their size, and as a result continue to decline in these grasslands, which are not commonly seen as habitats for threatened species. I wanted to do something that could help conserve this important and underappreciated biodiversity, and the effects of stock grazing became the big question for me.

Several months later, after encouragement from my biology teacher Remco Baars, I began seriously considering how I could gather data. I trialled a few different methods over November and December 2013, but each time I felt that the method was too convoluted, and the data that I received was not representative of what I had seen in each quadrat.

After those initial trials, and discussions with Mr Baars about potential improvements, I developed a method of data collection which I felt represented both the abiotic and biotic environment of each quadrat.

Method

The study sites were situated at two localities only 1–2km apart in an attempt to minimise additional variables created by changes in the physical environment. The ungrazed site, the Kōwhai Sanctuary, has been retired from grazing since 2008. The grazed site, Thompson’s Road, is grazed with a low stocking of sheep and cattle. It was stocked with 300 calves between July and October 2013. For the three months prior to my survey the site received no stock grazing, but growth is likely to have been minimal in the absence of grazing due to the progression of the summer drought.

To account for variation in micro-topography, stratified sampling was employed. Stratified sampling separates each site into four strata based on landform: terrace tops, channel floors, north facing scarps, and south facing scarps.

Randomly plotted 1x1m quadrats were surveyed, and soil depths were measured in the NE and SW diagonal corners, to produce a mean soil depth value. Measuring the soil depth allowed me to understand the abiotic environment of each quadrat to identify patterns in species composition, and answer my secondary aim.

To determine total surface cover, two 25x25cm frames, divided into five centimetre squares were placed in two corners of the quadrat and the following recorded: species present, and number of 5x5cm squares occupied per species.

This gathered a reliable representation of the species composition of a quadrat, and was repeated for 20 quadrats per study site; five randomly placed quadrats per strata type.

At this stage, I began to understand how restricted my knowledge of plant identification was, and that this was likely the biggest barrier to achieving my objective. I used publications such as What Grass Is That? by N.S. Lambrechtsen, and NZ Naturewatch to assist in identification. Colin Meurk was especially helpful with checking identifications. In the early days, while my knowledge was still poor, revelations, such as when I realised that I’d been recording three species of moss as one species, meant that I needed to resurvey a number of quadrats time and time again. I found that each quadrat could take up to an hour to survey, and by the end of my estimated 60 hours of field work I had become proficient in identifying over 40 different grassland species.

There was a 1–2 month period between starting and finishing surveying both sites. The progression of the summer drought over this period likely desiccated many exotic grasses, resulting in the abundant grass leaf litter at the  Kōwhai Sanctuary. To account for this additional variable I recorded grass litter and added it to its respective species per cent cover, however this compensation was conservative. While the drought reduced the accuracy of a comparison of exotic grass abundance between sites, it did not impact the comparison or native species. In hindsight, surveying of both sites should have been completed over a shorter time frame to reduce this variable.

Once I began collecting data, it became apparent that a multivariate analysis was necessary. I contacted Dr Susan Walker, a Landcare Research ecologist based in Dunedin, who agreed to help me complete my data analysis using ‘R’ programming. Susan and I worked to compile my raw data into Excel spreadsheets that could be accessed by ‘R’. Over the course of a few months, I spent hours on the phone with Susan, and then would practice the data script that we had covered – manipulating the generic script and data sheets to produce the desired graph. In this manner, I was able to generate multidimensional scaling (MDS) plots, boxplots, species accumulation curves, and double line graphs which explained trends in the data.

Results, including graphs

Figure 2 shows a gradient of plant communities ranging from low diversity exotic grasses to diverse herbaceous short turf dominated by native species, and exotic forbs and tufted grasses. This vegetation gradient, running from bottom right to top left, is believed to be caused by variations in soil depth. Grazed quadrats (red) and ungrazed quadrats (blue) were highly segregated, indicating that both sites have different species compositions.

Figure 3 investigates site segregation identified in the MDS plots. There was a large difference in the summed per cent cover of indigenous vascular species, which are virtually absent in the ungrazed site – a 24 per cent decrease in median summed per cent cover. Exotic vascular species showed a similar decrease of 26 per cent in median summed per cent cover.

Data from the grazed site and ungrazed site was incorporated into accumulation curves (Figure 4) to compare the species diversity between sites. Both native and exotic species show a decrease in diversity in the ungrazed site. In Figure 4a, the large separation of the two curves indicates that native vascular species incurred the largest loss in diversity. Figure 4b shows a decrease in the percentage of indigenous vascular species out of total species, consistent with diversity trends identified in (4a).

Soil depth was identified to be the most significant abiotic variable which makes it likely that changes in soil depth caused the diagonal distribution of quadrats and species, identified in the MDS plots (Figure 2). The bottom right of the MDS plots are typically characterised by deep soils (20cm–100cm) and exotic grasses on the MDS plots while the top left is characterised by thin soils (<20cm) and native species. While there were differences in the soil depth of some strata between sites, the mean soil depth of both sites only differs by 3.7cm.

Discussion

Comparing the abundance and diversity of species between sites shows that stock grazing enhances the survival of native prostrate species in the mid-Waimakariri floodplain, with the biggest effect occurring on deep soils (20+ cm), through the thinning of dense exotic grasses that would otherwise outcompete most native species.

Grazing is a form of disturbance which, like stress, generates low stature communities, often high in native species abundance and diversity. Sites which are highly stressed (soil depths less than 20cm), require fewer disturbances than deeper soils to maintain low stature vegetation. This is why shallow soils retained similar nonvascular and exotic species compositions irrespective of grazing treatment, however deeper soils changed considerably as a reduction in disturbance (grazing) allowed dense exotic grasses like cock’s foot (Dactylis glomerata) and red fescue (Festuca rubra) to develop on less stressed sites.

Conclusion

The results of this investigation show that, in semi-natural grasslands in the mid-Waimakariri floodplain, stock grazing can occur in conjunction with ecological conservation. However, since appropriate stocking rates and duration to promote native species were not investigated, the trends observed can only be applied in general terms. In the Harper and Avoca Catchment, North Canterbury, A. B. Rose etal. (1976) found that prolonged grazing “generally promoted decline in native species and invasion by exotic species...” It is expected that a similar trend would occur in the mid-Waimakariri floodplain through herbivory and trampling of palatable native species.

While this investigation indicates that stock grazing is beneficial to most remaining indigenous grassland species, grazing is highly detrimental for remnant shrubs, especially critically endangered Olearia adenocarpa, which was present in the grazed site but highly restricted, and generally grazed to ground level. Should semi-natural grasslands be grazed, areas of native shrubs should be fenced from stock to allow regeneration. The effects of agricultural intensification e.g. irrigation and fertilising, were not investigated; however, background reading suggested that it will reduce stress, therefore increasing competition from exotic species (Meurk & Hall, 2006), eventually displacing most native plants.

A large amount of New Zealand’s biodiversity is found as small grassland plants (Meurk & Greenep, 2002), yet herbaceous vegetation is declining at a rapid rate nationwide (Meurk et al., 2006), often as a consequence of agricultural intensification and weed invasions. However, the decision to reserve land for either “preservation or production” is not so black and white (H. Moller etal., 2008); rather sustainable land use through low-moderate intensity stock grazing could be considered to achieve both economic and ecologic goals without the need for compromise. Semi-natural grasslands in the mid-Waimakariri floodplain, and other similar ecosystems, are a unique opportunity to restore and conserve a biodiverse agricultural landscape.

In the future, perhaps as a university thesis, I’d like to undertake further research to identify an optimal stocking duration and density to promote native grassland species that can be easily incorporated into Environment Canterbury’s and landowners’ grazing regimes.

I encourage anyone, regardless of age, to get involved in science. Help is essential, so make use of the many scientists and experts nationwide who want nothing more than to help someone with a similar interest.

Acknowledgements

I thank George Mckay and Environment Canterbury who allowed me to work on their property, and Dr Colin Meurk from Landcare Research, who taught me species identification and identified any enigmas. A huge thanks goes to Dr Susan Walker, also from Landcare Research, who taught me how to use the statistics modelling program “R” and advised me on the representation of my data. Finally, I’d like to thank Remco Baars, head of science at Darfield High School, for guidance, revision, and motivation over the last year.