Gradients in edaphic features influence boundaries of mulga scrublands and hummock grasslands in Pilbara, Western Australia
Amelia C. Hawkins
Abstract
Mulga (Acacia aptaneura) shrub-land and hummock (Triodia) grassland boundaries are defined by both dynamic and static landscape features. This study explores the edaphic gradients across five study sites in the northern Hamersley Ranges of north-west Australia, and the limiting nutrients that underpin the mulga-Triodia boundaries found here. Species sampling of each site proved distinct taxonomic and structural composition within and between sites. This was compared with the physical and nutrient properties of soil samples taken from each site. The study found that Mulga shrub-land communities only occurred on N and P relatively rich soils, while tussock and hummock grassland communities occurred on overall nutrient poor soils. Although life-form diversity was low at both grassland sites, overall species diversity wasn’t limited by nutrient availability. The hummock grassland site had more than nine times less N (% wt) than the primary mulga shrub-land site, and more than three times less labile inorganic P (% wt). Despite this, both sites had close species diversity, site 1 (H’= 2.216), and site 5 (H’= 2.309). The hummock grassland site had high δC13 and C/N measures, suggesting high soil turnover of organic matter and microbial activity. The study found that there were no obvious environmental discontinuities demarcating the vegetation boundaries however the presence of A. aptaneura corresponded with higher overall macro-nutrients in soils. It is unclear whether this is a cause or effect.
Key words: soil nutrients, Pilbara, Mulga, shrublands, grasslands
1. Introduction
The high taxonomic and structural diversity of the Pilbara’s vegetation reflect the diverse geological and geomorphic diversity of the region (van Vreeswyk 2004). The Pilbara bioregion of Western Australia lies on the Archaean Pilbara craton (>3.5 bya) whose edaphic features are complex and by all accounts, deeply weathered (Kendrick 2001). Geology in the
Pilbara is predominantly Proterozoic meta-sediments that are overlain by Tertiary regolith (van Vreeswyk 2004). The region is flanked by two extensive sedimentary basins, the Canning Basin in the north-east, and the Canarvon Basin in the west. These create extensive groundwater reserves, supporting the pulse-reserve paradigm of the vegetation (Reeves 2007). Spanning 178, 060km2 of semi-arid land bisected by riparian arteries (BHP 2015), the region falls on a bioclimatic transitional zone, experiencing both tropical summer rain in lower latitudes, and mesic winter rain in higher latitudes (Kendrick 2001). The region is characterized by hummock and tussock grasslands, Acacia and occasional eucalypts, though recent floristic surveys count 2113 taxa in over 115 different families (Hamersley 2016). This anomalous diversity has classified the region one of Australia’s 15 biodiversity hotspots (Reeves 2007). Vegetation mosaics are pronounced due to complex layers of land tenure. The majority of the Pilbara is Unallocated Crown land, although various mining companies, pastoral lessees and the National Native Title Tribunal are all stakeholders with varying land management practices (Hamersley 2016). High disturbance, high geo-topographical heterogeneity and variable water supply are complex interplaying determinants of vegetation pattern across the region. This paper outlines the species present at five sites over the Munjina Claypan, their distribution, and the physico-chemical properties of the soils that underpin vegetation boundaries. Multivariate analyses are used to investigate the patterns of alpha and beta diversity across sites, comparing species composition and structure. Community data are compared with environmental data to better understand edaphic features corresponding to vegetation patterns, in particular the boundaries of mulga shrub-lands and Triodia grasslands.
2. Methods
2.1. Study area and experimental design
The study took place in July 2019 at the Munjina claypan (22’37’57S, 118’46’00E). The site is located in the northern part of the Hamersley Ranges in the Pilbara bioregion of north-west Australia (IBRA). The Hamersley Ranges vary topographically from hills, ridges and plateaux that bare isolated and continuous chains of emergent uplands (the Hamersley Ranges). These undulations support ephemeral freshwater wetlands and riparian tributaries from nearby Fortescue and Ashburton Rivers, and catchments (Lyons 2015). Surface geology is banded iron formation (BIF), with alluvium and colluvium forming the base of the valley floor soils. The Hamersley Ranges’ proximity to the coast is conducive to the formation of low pressure tropical cyclones, occurring from January to March. Thus, elements of both sub-
tropical and semi-arid zone vegetation are present (Kendrick 2001). Mean annual rainfall is ❤10mm (BOM 2018), and highly seasonal, with more than 80% of annual rainfall occurring between December and April. The average rainfall at the time the study was conducted (July 2019) was just 109mm, less than half of the average rainfall to July. Furthermore, 2018 studies recorded <200mm by July, thus the area has experienced a period of prolonged drought. Average diurnal temperatures range from 20°C to 40°C during the summer months and from 8°C to 28°C in the winter (BOM 2011).
Munjina
Claypan
Figure 1. Map of the Pilbara bioregion showing key locality of study, Munjina Claypan (Reeves et al. 2007)
Munjina Claypan was excised from Rio Tinto’s Juna Downs pastoral lease on July 1st 2015. It has since experienced varying management impacts on vegetation and landscape. The effects of which have been monitored over four years (2015–2019) of floristic surveys at five different sites. These sites vary in proximity to water-bores and swales, and subsequently, in vegetation composition. The sites are classified by their broad vegetation types and corresponding relief (Table 1).
Table 1. UTM (zone 50) co-ordinates of sampling plots across five different sites
Sites UTM co-ordinates Easting/Northing
Landscape position Broad veg type
1_1 671854 7483218 Flats Banded mulga
1_2 671936 7483298 Flats Banded mulga
1_3 672012 7483278 Flats Banded mulga
2_1 682217 7498060 Flats Mulga woodland
2_2 682384 7498208 Flats Mulga woodland
2_3 682289 7498096 Flats Mulga woodland
3_1 682438 7500165 Flats Mixed tussock grassland 3_2 682424 7500093 Flats Mixed tussock grassland 3_3 682430 7500250 Flats Mixed tussock grassland
4_1 666051 7480365 Hill top Hummock grassland/shrub steppe 4_2 665964 7480127 Hill top Hummock grassland/shrub steppe 4_3 666162 7480228 Hill top Hummock grassland/shrub steppe
5_1 669702 7479685 Lower slope Hummock grassland/shrub steppe 5_2 669621 7479712 Lower slope Hummock grassland/shrub steppe 5_3 669794 7479633 Lower slope Hummock grassland/shrub steppe
6_1 675818 7480467 Hilltop Mixed hummock
grassland/miscellaneous shrubs
Data was collected in 2019 from sites 1, 2, 3, and 5, however site 4 was burnt and unable to be surveyed. Site 4 is situated on a hilltop, and has distinct vegetation to the slopes and flats. To gauge taxonomic diversity of hilltop landscape position, we surveyed site 6. Only presence-absence data was collected at site 6 and as such, data was omitted for further analysis. Sites 1–3 occur on low relief, valley floor. Site 1 (banded mulga) is described as a sparse low woodland dominated by Acacia aptaneura and tussock grass. Site 2 (mulga woodland) is described as low open woodland dominated by A. aptaneura and Eucalyptus xerothermica. Site 3 (mixed tussock grassland) is described as an open tussock grassland dominated by Themeda triandra. Site 5 occurred on the lower slopes, and was decribed as open hummock grassland dominated by Triodia vanleeuwenii and T. pungens. Site classifications as defined by Carnahan, Specht and Beard & Webb are found in the appendix.
2.2. Sampling
In order to get data for species diversity and abundance, three 50m x 50m quadrats were established at each site (DEC standard). These quadrat sizes were reached progressively, starting with a 1m x 1m quadrat and doubling the sample area after a comprehensive species count was made for each quadrat size (1m2, 2m2, 4m2, 8m2, 16m2, 32m2, and 50m2). Abundance estimations were made using a modified Braun-Blanquet Method. This scale assigns each species a value based on its projective canopy cover, i.e. proportion of area that would be covered if you projected canopy cover on topographical view of the quadrat. Constructed species accumulation curves showed that all plots’ species accumulations plateaued by 2500m2 (Figure 2). Each species’ functional diversity traits were ranked using three criteria: life-form, Raunkiaer classification, and leaf-size (appendix).
2.3. Soil analysis
For each site, three soils samples (0–5 cm depth) were collected from different positions within the 50 m x 50 m plots, and then bulked by plot. Soil profiles were characterised using Northcote Soil Classification (1976). Soil samples were analysed at the West Australian Biogeochemistry Centre at The University of Western Australia. Samples were sieved (<2 mm) and then left to dry at 50°C before being pulverised and analysed for C, N, P and pH using the methods described in Grierson and Adams (2000). The percentage C and N of soils as well as the soil C and N isotope composition were analysed using a Sercon 20/22 isotope ratio mass spectrometer (Sercon, UK).
2.4. Analysis
2.4.1. Alpha diversity
Plant taxonomic richness, evenness and diversity were determined for each site, using pooled data from the sampling plots. These values were derived from abundance data collected from Braun-Blanquet cover assessments of each plot. Abundance data often have many zero values, thus the Shannon-Weiner diversity function was used to determine diversity values. Indices for richness (total species= S), evenness (Pielou’s evenness= J’) and concentration of dominance (Inverse-Simpson’s Index= 1-λ) were also generated for each plot. These indices were generated using the DIVERSE function in PRIMER 6.0 (Clarke 2015). Species rank abundance (log) curves were produced to underscore the concentration of dominance of a select few species at each site. This is computed using Inverse-Simpson’s Index, which yields the probability that two individuals drawn at random from a sample will be different species. In order to determine plot sampling size sufficiency, species accumulation curves were
generated using the charting function in Microsoft Excel. Logarithmic curves were prepared for each plot at each site. Adequate sampling size is represented by a plateau in species accumulation at maximum plot size. Further using the charting function on Excel, a stacked column graph was prepared, showing life-form diversity as a function of Braun-Blanquet (BB) cover for each plot.
2.4.2. Beta diversity
Multivariate analyses were carried out using PRIMER (Clarke 2015), in order to resolve species turn-over between sites. Abundance data were square-root transformed for all multivariate functions performed to reduce the weight of large values produced as a function of BB cover. The transformed data were used to generate Bray-Curtis resemblance matrices. Non-metric multidimensional scaling ordination was produced to display spatial relatedness or lack thereof sites’ composition. Similarity clusters were superimposed to complement this, with resemblance levels at 20, 40, 60 and 80. ANOSIM (Analysis of Similarity) was used to determine whether these observed differences in site species composition had statistical significance. ANOSIM uses non-parametric permutation procedures applied to similarity matrices based on rank similarities between samples. The ANOSIM function yields an R statistic value between -1 to 1. Values more proximate to 1 indicate a large difference between groups. Values proximate to 0 indicate little or negligible difference between groups. The SIMPER (Similarity of Percentages) function was performed next to show species that were the greatest contributors to (Bray-Curtis) dissimilarities between sites.
2.4.3. Environmental determinants
Data on environmental factors were collected and collated, to compare between sites. Environmental determinants are often inextricably linked, thus a PCA (Principal Component Analysis) was performed using PRIMER to display relativeness of factors to one another and their relativeness to each site. PCA output shows how much a variable (environmental factor) contributes to calculating the principal component on either axis. If the variable’s coefficient has strong associations with either principal component, then the factor is largely explained by it. To support environmental-community correlations, a BIO-ENV procedure was performed. The results list the most correlative variables for explaining cover data nMDS. ANOSIM was used to then infer statistical significance of environmental-community correlates.
3. Results
3.1. Alpha diversity
Plant species richness, evenness, and diversity were compared across sites 1, 2, 3, and 5. In total, 281 species were recorded from 65 different genera, the richest being Acacia (26 spp) and Aristida (26 spp). Mean species richness was highest at site 1, with 27 species observed, and site 5, with 27 species. However, both sites had disproportionately high concentrations of species richness in just one plot at each site. Plot 1_3 at site 1 had at least 20 more species than both other plots at that site. Similarly, plot 5_2 at site 5 had at least 10 more species than both other plots (Table 2). Mean species evenness (Pielou’s Index) was greatest at site 5 (J’= 0.699). Site 1 had the second greatest value (J’= 0.682), followed closely by site 2 (J’= 0.069). Shannon’s diversity index showed that mean diversity was highest at site 5 (H’= 2.03) and site 1 (H’= 2.216), and lowest at site 3 (H’= 1.280). A more even distribution of species composition is observed at site 5, with a mean inverse Simpson’s dominance Index of 0.821, followed closely by site 1 with a mean Index of 0.809. Site 3 had low inverse Simpson’s indices across all plots, with a mean index of 0.511 (Table 2). This indicates that the site was highly dominated by a small number of abundant species. This dominance is visible as a species rank abundance curve for each sampling plot at each site (Figure 2). This figure shows that >60% of abundance at site 3 (mixed tussock grassland) was dominated by just a few species. Species accumulation curves for sampling plots at each site suggest that plot size was sufficient for obtaining values that reflect the true alpha-diversity at each site (Figure 3). The species-accumulation curve for site 5_2 however was steeper than other plots, indicating new species were being found at the 50m x 50m site, and sampling size was insufficient in order to capture true diversity.
Table 2. Species richness, evenness and diversity for study sites 1–3 and 5.
Sites Total species richness
Pielou’s
evenness Index (J’)
Shannon’s
Diversity Index (H’ log-e)
Inverse Simpson’s Index (1-ʎ)
1_1 17 0.707 2.002 0.785 1_2 22 0.552 1.707 0.719 1_3 42 0.786 2.938 0.924 µ, ± 27 ± 10.801 0.682 ±0.097 2.216 ± 0.525 0.809 ±0.085
2_1 19 0.772 2.272 0.835
2_2 21 0.552 1.679 0.687
2_3 17 0.685 1.939 0.787
µ, ± 19 ± 1.633 0.669 ± 0.092 1.964 ± 0.243 0.769 ± 0.062
3_1 13 0.517 1.325 0.493
3_2 12 0.509 1.266 0.570
3_3 16 0.451 1.249 0.471
µ, ± 14 ± 1.699 0.492 ± 0.029 1.280 ± 0.032 0.511 ± 0.043
5_1 24 0.686 2.179 0.777
5_2 35 0.764 2.715 0.892
5_3 23 0.649 2.034 0.794
µ, ± 27 ± 5.437 0.699 ± 0.048 2.309 ± 0.292 0.821 ± 0.507
Figure 2. Species rank abundance curves (log-transformed) for each site, all species included.
s
e
i
c
e
p
s
f
o
#
50
Site 5, plot 2
40
30
20
10
0
0 500 1000 1500 2000 2500 3000
-10
Area m2
Figure 3. Species-accumulation curves for each plot, log-transformed. Green= site 5; Red= site 1; Blue= site 2; Black= site 3.
Relative abundance of life-form as a percentage of cover at each site may be visualised in Figure 4. Cover at site 1 is >50% tree canopy, attributed predominantly to A. aptaneura, and E. xerothermica. Another ~25% of cover is tussock grass, mostly represented by T.a triandra. Site 2 has >50% tree cover attributed to A. aptaneura. The remainder cover is largely represented by T. triandra. Site 3 is >98% tussock dominated. T. triandra and
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1 2 3 5
herb hummock grass shrub <1 shrub 1–2 tree (canopy) tree (sub-dominant) tussock grass
Aristida latifolia are the sites two biggest representatives. Finally, site 5 is >80% hummock grass, of which T. vanleuwenii and T. pungens comprise the largest proportion. Other prevalent shrubs (>1m<2m) at site 5 belonged almost exclusively to Acacia.
Figure 4. Stacked column plot (scaled to 100%) of diversity of life-form at each site, using Braun-Blanquet cover data (shrub measuring unit is in metres).
3.2. Beta diversity: spatial turnover in vegetation types
Figure 5. Non-metric multidimensional scaling ordination for all study sites, each point represents a sampling plot. Similarity defined by species abundance as a function of Braun Blanquet cover. Data has been square-root transformed.
Table 3. SIMPER output showing species that contribute most to dissimilarity between sites SITE Discriminating species Contribution to dissimilarity (%)
Banded Mulga (1)
Mulga woodland (2)
Mixed tussock grassland
(3)
Hummock
grasslands/shrub steppe
(5)
Acacia aptaneura
Eucalyptus xerothermica Hakea lorea subsp. lorea Acacia aptaneura
Themeda triandra
Sida fibulifera
Themeda triandra
Aristida latifolia
Altenanthera nana
Triodia pungens
Triodia vanleeuwenii Acacia tenuissima
40.9 32.42 13.18 56.38 19.54 3.53
85.45 2.65 2.42 36.52 16.82 13.91
A high rate of species turnover was recorded between sites (Figure 5). Sampling plots at site 3 had very high similarity values, and were compositionally very close (Table 3). Sampling plots 1_2, and 1_1 showed more similarity to plots 2_3 and 2_2, while plot 2_1 and 1_3 were less similar. This is likely on account of these sampling plots’ proximity to water bore/swales. ANOSIM global tests confirmed that each sampling plot was significantly compositionally different to each other (R= 0.361, P< 0.037). Pairwise tests revealed that site 3 (mixed tussock grassland) was the most different site to every other site. Site 1 and site 3 pairing had an R value of 0.556 and a p-value of <0.01. Site 2 and site 3 pairing had an R value of 0.889 and a p-value of <0.01.
SIMPER analysis resolved which species contributed most to the dissimilarities between sites (Table 3). The most discriminating species for site 1 were A. aptaneura and E. xerothermica, contributing >70% dissimilarity between them. Accounting for >75% dissimilarity rate at site 2 were A. aptaneura and T. triandra. At site 3, contributing >85% dissimilarity was the same T. triandra. At site 5, T. pungens and T. vanleeuwenii were the greatest contributors (>50%).
3.3. Soil profile analyses
ANOSIM results showed that there was a significant difference (R= 0.657, p<0.01) in nutrient profiles of soils across sites. Results from the BEST (BIO-ENV) function in primer
elucidate which of these environmental correlates best explain patterns in community data (Table 4). Variable combinations that best explained these patterns were are seen in table 4.
Table 4. BEST (BIO-ENV) results from PRIMER showing which environmental variables best explain community patterns
Number of variables Correlation Variable combinations 4 0.739 Elevation, δ15N, δ13C, C/N 5 0.713 Elevation, δ15N, δ13C, C/N, OH labile inorganic P
5 0.702 Elevation, pH, δ15N, δ13C
Mean δ15N values were highest in sites 1 (= 6.56‰ AIR) and 2 (= 6.94‰ AIR). Nitrogen measurements by percentage weight were also highest in these two sites, site 1 (=0.25%) and site 2 (=0.143%). δ13C values were highest at site 3 (=27.458‰ VPDB) and site 5 (=26.328‰ VPDB). Carbon by percentage weight however was much lower at these sites. Carbon and nitrogen ratio values were comparable across sites, with site 5 having the largest value (=19.048) and site 3 having the lowest value (=12.599). Labile inorganic P was highest at site 2 (=42.084) followed closely by site 1 (=33.13). Site 5 was particularly P deficient with just 9.449 ug/g labile organic P measured. Soil acidity varied across sites. Site 5 was the most acidic with a pH level of 5.097. Site 1 had the least acidic soils, with a pH of 6.47 (Table 5).
Table 5. Mean and standard deviation for soil nutrients at each site (both δ15N (AIR standard) and δ13C (VPDB standard) values are per mil)
Sites pHwater 1:1
δ15N δ13C N [wt %]
C [wt %]
C/N OH labile inorganic
P (ug/g)
OH
extractable total P
(ug/g)
1_1 7.39 5.57 21.87 0.08 1.3 16.6 12.37 30.7 1_2 6.29 6.91 21.22 0.06 0.9 15.3 42.61 73.8 1_3 5.73 7.19 24.42 0.60 6.20 10.30 44.42 197.7
µ, ± 6.47 ± 0.689
6.56 ± 0.709
22.50 ± 1.381
0.25 ± 0.250
2.81 ± 2.404
14.05 ± 2.706
33.13 ± 14.7
100.71 ± 70.787
2_1 5.5 7.32 23.60 0.23 3.10 13.48 51.77 125.9
2_2 5.85 6.32 22.64 0.07 1.1 14.7 30.00 58.6 2_3 5.71 7.18 21.38 0.12 2.3 18.4 44.48 85.6
µ, ± 5.687 ± 0.144
6.940 ± 0.443
22.541 ± 0.909
0.143 ± 0.065
2.161 ±
0.828
15.524 ± 2.093
42.084 ± 9.046
90.05 ± 27.64
3_1 6.07 5.94 28.26 0.05 0.6 13.1 19.97 41.8 3_2 6.81 6.12 25.73 0.05 0.6 12.2 25.77 37.4 3_3 6.36 6.60 28.38 0.05 0.6 12.5 28.15 46.0
µ, ± 6.413 ± 0.304
6.22 ± 0.277
27.458 ± 1.225
0.049 ± 0.001
0.623 ±
0.023
12.599 ± 0.358
24.63 ± 3.438
41.729 ± 3.508
5_1 5.43 4.72 26.85 0.02 0.5 22.4 9.35 33.7 5_2 4.71 4.54 26.61 0.04 0.5 14.9 5.15 30.7 5_3 5.15 3.94 25.53 0.02 0.4 19.9 13.85 31.0
µ, ± 5.097 ± 0.297
4.401 ±0.334
26.328 ± 0.574
0.027 ± 0.007
0.486 ±
0.039
19.048 ± 3.111
9.449 ± 3.551
31.79 ± 1.341
Ordination of community data with PCA (Principal Component Analysis) showed that the first three PC axes accounted cumulatively for 85.7% of the total species variance (Figure 6). The first PC axis accounted for 51.4% of the total variance and the second axis for 20.4%. The first component had strong negative associations with variables δ15N, OH labile inorganic P, OH extractable total P, C and N. The second component had strong negative associations with elevation, N and C. Figure 6 shows that site 1_3 is obviously aberrant, associated with variables OH extractable total P, C and N (similarity cluster at level 80). All plots at site 5 are tightly clustered having associations with C/N ratio and elevation. Site 2_1 appears to be closely linked with δ15N.
Figure 6. PCA plot showing variable association to PC1 and PC2
4. Discussion
4.1. Hydrology and Pulse-Reserve paradigms
This study is intended to be used referentially with previous years’ study of the Munjina Claypan since its decommission from the Juna Downs Station. Little evidence of disturbance from grazing, soil compaction or fire (>10 years unburnt) was found across sites, although drought disturbance was a major factor affecting species diversity and distribution in 2019. Clear boundaries are found between sites that harbour mulga and other woody shrubs, as well as boundaries of hummock and tussock dominated grasslands.
Our study site had experienced a prolonged period of drought at the time of sampling (BOM 2019), though the region is characterised by variable highly stochastic water supply both temporally and spatially (Kendrick 2001). Vegetation is thus adapted to highly seasonal rainfall occurring in ‘pulses’, and prolonged periods of little to no rainfall, during which water supply is in deep ‘reserves’. Rainfall events promote mineralisation of organic and mineral matter in soils (Cowling 1996) thus promoting plant growth and establishment. The clay-loam soils of the study site are well suited to this pulse-reserve model as clay has low permeation rates (Kneeshaw 2004) but high rates of water retention. High succession ‘pulse’
rainfall events recharge ‘reserves’, important water sources for Pilbara vegetation (Cowling 1996). The study took place during a ‘reserve’ phase, and as such there was lower nutrient bioavailability (Ingram 2001).
4.2. Diversity in nutrient deficient soils
At a regional scale (10km to 106km2), high diversity often occurs in frequently fire-disturbed and nutrient poor soils (Cowling 1996). However, at a local scale (<10km2) it appears that other factors are at play. The study found that higher soil macro-nutrients supported greater life-form diversity, particularly woody mulga shrub-land species. It is unclear whether this is a cause or effect of their presence. Sites 1 (mulga shrub-land) and 5 (hummock-grassland) were the two most diverse sites. Life-form diversity however was low at site 5, dominated by many species of hummock. Life-form diversity was higher at site 1, which had a mixture of trees, tussocks and shrubs (both <1m and >1m<2m). It is likely that species turnover between the two sites was underpinned by nutrient content and water supply rather than soil physical properties. Site 1 had higher P, N and C soil content, as well higher soil pH, likely explaining the ability to support greater life-form diversity (Nano 2008). Nano and Clarke 2008 have suggested that turnover of floristic and structural features between mulga shrub-lands and hummock grasslands are defined by differing clay content, pH and soil depth. Clay content and soil depth determine soil water retention, and pH regulates nutrient availability through red-ox potential and solubility equilibria (Brundrett 2002). Our study showed that neither site differed in soil depth (both classed as medium) or clay content (both classed as clay-loam). Site 5 had relatively acidic soils (pH= 5.097) which may have affected bioavailability of nutrients. Furthermore, site 5 had the greatest C/N ratio, an important limitation on the decomposability of organic material in soil (Zhou 2019). Hummock grasslands are known for having nitrogen-depleting microbial associations with cyanobacteria, lichen and moss (Elridge 2001), which may explain this C/N ratio.
The different leaf-area index (LAI) of these two sites is indicative of water availability (O’Grady 2011). Mulga shrub-land vegetation will typically have larger LAI’s than that of hummock grasslands (Morton 2011). This can be attributed to the landscape relief of each site. Site 1 occurs on the valley floor and as such is a drainage site, while site 5 occurs on a slight slope, where there is greater water run-off.
Sites 1 and 2 were closely related in species composition, likely due to their similar nutrient
profile. Greater presence of Acacia at both likely created a positive feedback, as Acacia are part of Fabaceae, known for harbouring nitrogen-fixing bacteria, thus greater soil N input. The slightly higher nitrogen content at site 1 is probably due to plots being located on a transect near a dismantled watering point, where a strong piosphere effect is observed. Plot 1_3 was the closest to the watering point and this is notable in its disproportionate species diversity and high relative nutrient content. Locales close to watering points will attract more fauna, thus more nitrogen rich excreta. Since it is a decommissioned site (>5 years ago), there was little trace of bovine presence or grazing impact. The enriched δ15N at the mulga shrub land sites (1 and 2) compared with the hummock grassland is related to the total N content, as there is a clear relationship between the two. High levels of soil δ15N are normally associated with quick soil turnover and loss of N, however there is an obvious feedback between the N fixation at the mulga sites and the loss of N in cycling (Cook 2005). Lambers 2008 highlights the eco-system level impact of P-impoverished soils on total N content. Limited P constrains the assimilation of N into plant biomass, such that biomass and necromass are both N and P limited (Lambers 2008). This negative feedback may be observed in N and P impoverished sites 3 and 5.
Sites 1 and 2 had the highest levels of carbon by percentage weight and phosphorus in labile inorganic form. These sites receive greater carbon and phosphorus-rich leaf litter than the grass-dominated sites 3 and 5 (BHP 2015) on account of higher overall LAI. Higher levels of phosphorus may also be associated with root adaptations to phosphorus-poor soils which exude organic acids and extracellular enzymes, capable of solubilising and mineralising otherwise metal-bound phosphorus (Lambers 2010).
The low diversity index of site 3 may be explained by dormancy of microbial and mycorrhizal associations due to drought (Brundrett 2002). Microbial associations in the rhizosphere are extremely important for the turnover of organic material and leaf litter, the major source of N, C and P for tussock grass. During times of drought, these microbes may enter a state of dormancy until hydrated, thus nutrient bioavailability in the rhizosphere is extremely low and only highly adapted plants (T. triandra) can establish themselves. Higher measures of δ13C in soils at site 3 and 5 are reflections of the C4 dominated communities. C4 grasses have less isotopic discrimination than C3 plants (Bowman 2007) and as such are good indicators of changing ecotones between these C3 and C4 communities. Because the
13C/12C ratio of soil organic matter is roughly equal to the 13C/12C of the plant from which its derived, changing 13C content in soil signifies changing vegetation, i.e. from C3 dominant to C4 dominant. This method for measuring soil organic matter turnover is important for monitoring vegetation boundaries and changes in those boundaries (Balesdent 1987). Understanding changes in vegetation composition and distribution over both temporal and spatial scales allows us to distinguish changes attributable to environmental factors from those of management and disturbance regimes.
Acknowledgements
I would like to thank Pauline Grierson and Michael Renton, who organised logistically and otherwise the specificities of the study site since its excision from Juna Station in 2015. Both Pauline and Michael have been invaluable mentors and inexhaustible sources of knowledge. Alison O’Donnell, Hannah Etchells, Doug Ford and Neil Pettit were amazing field support and subsequent species identification help. Kate Bowler sustained everybody on the field trip with her warm hospitality and camp management. This paper contributes to the final component of the specialist botany unit Australian Vegetation (PLNT3306) with the University of Western Australia.
References
Balesdent, Mariotti Andre, Guillet Bernard (1987) Natural 13C abundance as a tracer for studies of soil organic matter dynamics. Soil Biology and Biochemistry, 19(1), 25–30.
Barton P. S., Manning A. D., Gibb H. et al. (2009) Conserving ground-dwelling beetles in an endangered woodland community: multi-scale habitat effects on assemblage diversity. Biol. Conserv. 142, 1701–9.
BHP Biliton Iron Ore, Golder Associates Pty Ltd (2015). Ecohydrological conceptualisation of the Marillana Creek Region. Environmental Impact Assessment report. Report no.: 137646040–002-R-Rev4
Bowman, D., Boggs, G., Prior, L., & Krull, E. (2007). Dynamics of Acacia aneura-Triodia boundaries using carbon (14C and δ13C) and nitrogen (δ15N) signatures in soil organic matter in central Australia. Holocene, 17(3), 311–318.
Briggs, J. M, Knapp A, Blair, J.M, Heisler, L. J, Hoch, G. A, Lett, M.S, McCarron, J.K. (2005). An Ecosystem in Transition: Causes and Consequences of the Conversion of Mesic Grassland to Shrubland, BioScience, 55(3), 243–254
Brundrett M.C. (2002). Coevolution of roots and mycorrhizas of land plants. New Phytologist, 154: 275–304.
Clarke, K.R. and Gorley, R.N. (2015) PRIMER v7: User Manual/Tutorial. PRIMER EPlymouth.
Cowling, Richard & Rundel, Philip & Lamont, Byron & Kalin, Mary & Arianoutsou, Margarita. (1996). Plant diversity in Mediterranean-climate regions. Trends in ecology & evolution. 11, 362–6.
Environment Australia (2000). Revision of the Interim Biogeographic Regionalisation for Australia (IBRA) and Development of Version 5.1. Summary report Nov 2000.
Zhou Gyiang, Shan Xu, Philippe Ciais, Stefano Manzoni, Jingyun Fang, Guirui Yu, Xuli Tang, Ping Zhou, Wantong Wang, Junhua Yan, Gengxu Wang, Keping Ma, Shenggong Li, Sheng Du, Shijie Han, Youxin Ma, Deqiang Zhang, Juxiu Liu, Shizhong Liu, Guowei Chu, Qianmei Zhang, Yuelin Li, Wenjuan Huang, Hai Ren, Xiankai Lu, Xiuzhi Chen (2019). Climate and litter C/N ratio constrain soil organic carbon accumulation, National Science Review, 6(4), 746–757
Hamersley Iron Pty Ltd, Rio Tinto (2016). Flora Vegetation and Fauna Habitat Assessment at Juna Downs. Native Vegetation Clearing Permit- Supporting report.
Ingram, L (2001). Growth, nutrient cycling, and grazing of three perennial tussock grasses of the Pilbara region of NW Australia. Doctor of Philosophy, University of Western Australia.
John M. Briggs, Alan K. Knapp, John M. Blair, Jana L. Heisler, Greg A. Hoch, Michelle S. Lett, James K. McCarron. (2005) An Ecosystem in Transition: Causes and Consequences of the Conversion of Mesic Grassland to Shrubland, BioScience, 55(3), 243–254.
Kendrick, P. (2001). Pilbara 3 (PIL3 — Hamersley synopsis). Pp 568–580, in May, J.E. and N.L. McKenzie (2003). A Biodiversity Audit of Western Australia’s Biogeographical Subregions in 2002. Department of Conservation and Land Management, Western Australia.
Lambers, Hans & Brundrett, Mark & Raven, John & Hopper, Stephen. (2010). Plant mineral nutrition in ancient landscapes: High plant species diversity on infertile soils is linked to functional diversity for nutritional strategies. Plant and Soil. 348. 7–27
Lewandrowski, W. (2016). An ecophysiological approach to understanding recruitment in keystone Triodia species in arid zone restoration. The University of Western Australia.
Lyons, M. N. (2015). The riparian flora and plant communities of the Pilbara region of Western Australia. Records of the Western Australian Museum, Supplement, 78, 485–513.
Morton, S. R., Stafford Smith, D. M., Dickman, C. R., Dunkerley, D. L., Friedel, M. H., McAllister, R. R. J.,Westoby, M. (2011). A fresh framework for the ecology of arid Australia. Journal of Arid Environments, 75(4), 313–329.
Nano, C. E. M., Pavey, C. R (2013). Refining the ‘pulse-reserve’ model for arid central Australia: Seasonal rainfall, soil moisture and plant productivity in sand ridge and stony plain habitats of the Simpson Desert. Austral Ecology 38, 741–753.
Nano, C.E.M. & Clarke, P.J (2010). Woody-grass ratios in a grassy arid system are limited by multi-causal interactions of abiotic constraint, competition and fire. Oecologia. 162, 719.
Nicholas A.M, Franklin D.C, Bowman D.M.J.S (2009) Coexistence of shrubs and grass in a semi-arid landscape: a case study of mulga (Acacia aneura, Mimosaceae) shrublands embedded in fire-prone spinifex (Triodia pungens, Poaceae) hummock grasslands. Australian Journal of Botany 57, 396–405.
Nicholas, A. M (2007) Characteristics and dynamics of the mulga-spinifex boundaries at Mt. Denison Station in Central Australia. Student Thesis: Masters by Research- CDU.
O’Grady, A.P, Carter, J.L. & Bruce, J. (2011). Can we predict groundwater discharge from terrestrial ecosystems using eco-hydrological principals? Hydrol. Earth Syst. Sci. Discuss. 8, 8231–8253
Payne, A.L., A.A. Mitchell and W.F. Holman (1988). An inventory and condition survey of rangelands in the Ashburton River catchment, Western Australia. Department of Agriculture Technical Bulletin №62, revised edition 1988.
Reeves, Jessica & De, Patrick & Ae, Deckker & Halse, Stuart (2007). Groundwater Ostracods from the arid Pilbara region of northwestern Australia: Distribution and water chemistry. Hydrobiologia. 585, 99–118.
Van Leeuwen, S. and B. Bromilow. (2002). Botanical Survey of Hamersley Range Uplands. National Reserve System Project N709 Final Report — May 2002. Unpublished report prepared by the Department of Conservation and Land Management, Science Division.
Van Vreeswyk, A M, Leighton, K A, Payne, A L, and Hennig, P. (2004), An inventory and condition survey of the Pilbara region, Western Australia. Department of Agriculture and Food, Western Australia, Perth. Technical Bulletin 92.
Appendix
1. Vegetation classification schemes of study sites
Table 1. Study sites and vegetation classification scheme as per the NVIS (National Vegetation Information System)
Site 1 Site 2 Site 3 Site 5 Site 6
Vegetation Class
Shrub-land Woodland Tussock grass Hummock grass
Miscellaneous forbs/tussock grass
Structural Formation
Sparse low shrub-land Open low woodland
Open tussock grassland
Open
hummock grassland
Sparse
grassland/open herbfield
Broad
Floristic Formation
Sub
formation
Acacia aptaneura woody shrubs,
Isolated trees,
Eucalyptus/Corymbia; isolated shrubs,
Acacia; Themeda grass
Acacia
aptaneura low open woodland
Acacia low open
woodland;
Themeda grass
Themeda
triandra open grassland
Themeda
grassland,
ephemeral
forbs, isolated Eucalyptus
Triodia open hummock
grassland
Triodia open hummock
grassland,
Acacia small isolated shrubs.
P/A data for this site only, no cover data
P/A data for this site only, no cover data
Association U: Eucalyptus xerothermica
M: Acacia aptaneura
G: Cenchrus ciliaris,
Chyrsopogon falax,
Hakea lorea, sbspp
lorea.
U: Eucalyptus xerothermica, Eucalyptus socialis
M: Eragrostis, Themeda
U: Themeda triandra,
Aristida
Latifolia,
U: Eucalyptus socialis, E. xerothermica, A. aptaneura
M: Triodia vanleeuwenii, Triodia
pungens,
Triodia
melvileii
Table 2. Carnahan classification scheme
Sites Carnahan class
1 wS2yx (Acacia tall shrubs, foliage cover >10%❤0%, tall open shrub-land with tussock and misc. forb understory.
2 wM2ey (Acacia medium tree, foliage cover >10%❤0%, tall open shrub-land with tussock understory and associated Eucalyptus.
3 yG4a (Triandra tussock grassland, foliage cover >70%, dry open grassland with associated Astrebla)
5 tH4wx (Triodia hummock grassland, foliage cover >70%, dry open grassland with associated Acacia shrubs and misc. forbs.
6 Site unassessed for cover, P/A only. Mixed herbaceous and tussock grass with scattered Eucalyptus.
2. Species list
Species name Life Form Leaf Size Raunkiaer Weed or native
Abitulon dioicum shrub <1 microphyll nanophanaerophyte native Abutilon fraseri shrub <1 microphyll nanophanaerophyte native Abutilon lepidum shrub <1 microphyll nanophanaerophyte native Abutilon otocarpum shrub <1 microphyll nanophanaerophyte native Acacia adoxa var. adoxa shrub <1 microphyll nanophanaerophyte native Acacia ampliceps x sclerosperma shrub <1 microphyll nanophanaerophyte native Acacia aptaneura tree nanophyll microphanaerophyte native Acacia colei var. colei shrub 1–2 microphyll nanophanaerophyte native Acacia dictyophleba shrub 1–2 microphyll nanophanaerophyte native Acacia inaequilatera shrub 1–2 microphyll nanophanaerophyte native Acacia monticola shrub <1 nanophyll nanophanaerophyte native Acacia pachyacra shrub 1–2 microphyll nanophanaerophyte native Acacia pyrifolia shrub 1–2 microphyll nanophanaerophyte native Acacia rhodophloia x sibirica shrub 1–2 microphyll nanophanaerophyte native Acacia sibirica shrub 1–2 microphyll nanophanaerophyte native Acacia tenuissima shrub 1–2 leptophyll nanophanaerophyte native Acacia tetragonaphylla shrub 1–2 leptophyll nanophanaerophyte native Acacia tumida var. pilbarensis shrub 1–2 leptophyll nanophanaerophyte native Acacia victoriae shrub <1 microphyll nanophanaerophyte native Achyranthes aspera herb microphyll nanophanaerophyte native Altenanthera nana herb nanophyll hemicryptophyte native Amaranthus undulatus herb nanophyll hemicryptophyte native Amaranthus” green tooth” herb nanophyll hemicryptophyte native Anthobolus leptomerioides shrub <1 leafless nanophanaerophyte native Aristida contorta tussock grass nanophyll therophyte native Aristida holathera tussock grass nanophyll therophyte native Aristida inaequiglumis tussock grass nanophyll therophyte native Aristida latifolia tussock grass nanophyll therophyte native Bidens bipinnata herb microphyll therophyte weed Bothriochloa ewartiana tussock grass nanophyll therophyte native Calandrinia ptychosperma herb leptophyll therophyte native
Capparis lasiantha shrub 1–2 microphyll nanophanaerophyte native Cenchrus ciliaris tussock grass nanophyll therophyte native Cenchrus setiger tussock grass nanophyll therophyte native Cheilanthes sieberi fern leptophyll hemicryptophyte native Chloris pectinata tussock grass nanophyll therophyte native Chrysopogon fallax tussock grass nanophyll therophyte native Cleome viscosa herb leptophyll therophyte native Corchorus lasiocarpus shrub <1 microphyll nanophanaerophyte native Corymbia hamersleyana tree (canopy) nanophyll mesophanerophyte native Cymbopogon ambiguus tussock grass leptophyll therophyte native Cymbopogon obtectus tussock grass leptophyll therophyte native Dactyloctenium radulans tussock grass microphyll therophyte native Dampiera candicans shrub <1 microphyll nanophanaerophyte native Dichanthium sericeum tussock grass microphyll therophyte native Duppereya conmixta creeper microphyll chamaephyte native Dysphania kalpari herb leptophyll hemicryptophyte native Dysphania melanocarpa herb leptophyll hemicryptophyte native Enneapogon polyphyllus tussock grass nanophyll therophyte native Eragrostis cummingii tussock grass nanophyll therophyte native Eragrostis desertorum tussock grass nanophyll therophyte native Eragrostis eriopoda tussock grass nanophyll therophyte native Eragrostis xerophila tussock grass nanophyll therophyte native Eremophila lanceolata shrub <1 microphyll nanophanaerophyte native Eremophila longifolia shrub 1–2 microphyll nanophanaerophyte native Eriachnae benthamii tussock grass nanophyll therophyte native Eriachne mucronata tussock grass nanophyll therophyte native Eucalyptus gammophyla tree (sub-dominant) mesophyll microphanerophyte native Eucalyptus leucopholia tree (sub-dominant) mesophyll microphanerophyte native Eucalyptus socialis tree (sub-dominant) mesophyll microphanerophyte native Eucalyptus xerothermica tree mesophyll microphanerophyte native Eulalia aurea herb leptophyll hemicryptophyte native
Euphorbia tannensis subsp. eremophila
herb leptophyll hemicryptophyte native
Evolvulus alsinoides var. villosicalyx herb leptophyll hemicryptophyte native G269 herb leptophyll hemicryptophyte native G3sp92 herb leptophyll hemicryptophyte native Glycine falcata herb leptophyll hemicryptophyte native Gompholobium karijini shrub <1 microphyll nanophanaerophyte native
Gomphrena affinis subsp. pilbarensis
herb leptophyll therophyte native
Goodenia microptera herb microphyll hemicryptophyte native Goodenia stobbsiana herb microphyll hemicryptophyte native Goodenia triodiophyla herb microphyll hemicryptophyte native
Gossypium robinsonii shrub 1–2 microphyll nanophanaerophyte native Grevillea wickhamii shrub>2 microphyll microphanerophyte native Hakea chordophylla shrub>2 nanophyll microphanerophyte native Hakea lorea subsp. lorea tree (sub-dominant) nanophyll microphanerophyte native Hibiscus sturtii shrub 1–2 microphyll nanophanaerophyte native Indigofera boviperda shrub <1 microphyll nanophanaerophyte native Indigofera georgei shrub <1 microphyll nanophanaerophyte native Indigofera monophylla shrub <1 microphyll nanophanaerophyte native Iseilema membranaeceum tussock grass microphyll therophyte native jasminium didymum creeper microphyll hemicryptophyte native Maeriana villosa shrub <1 leptophyll nanophanaerophyte native Malvastrum americanum herb nanophyll hemicryptophyte weed Mistletoe parasite nanophyll unknown weed
Paspalidium rarum tussock grass microphyll therophyte native Petalostylis labicheoides shrub 1–2 nanophyll nanophanaerophyte native Pluchea dentex herb nanophyll hemicryptophyte n
Polycarpaea corymbosa var. corymbosa
herb leptophyll therophyte native
Polygala sp.prostrate herb nanophyll therophyte native Polymeria longifolia herb nanophyll therophyte native Portulaca oleracea herb nanophyll therophyte native Pterocaulon sphacelatum herb microphyll therophyte native Pterocaulon sphaeranthoides herb microphyll therophyte native
Ptilotus calostachyus var calostachyus
herb leptophyll therophyte native
Ptilotus obovatus shrub <1 nanophyll nanophanaerophyte native Ptilotus rotundifolius shrub <1 nanophyll nanophanaerophyte native Rhagodia eramaea shrub <1 nanophyll nanophanaerophyte native Rhynchosia minima creeper nanophyll hemicryptophyte native Salsola australis shrub <1 leptophyll chamaephyte native Scaevola paviflora shrub <1 leptophyll chamaephyte native Sclerolaena cornishiana shrub <1 leptophyll chamaephyte native
Senna artemisioides subsp. artemisioides
shrub <1 microphyll nanophanaerophyte native
Senna artemisioides subsp. ficifolia shrub <1 microphyll nanophanaerophyte native Senna glutinosa subsp. glutinosa shrub <1 microphyll nanophanaerophyte native Senna glutinosa subsp. pruinosa shrub <1 microphyll nanophanaerophyte native Senna notabilis shrub <1 microphyll nanophanaerophyte native Senna pleurocarpa var. angustifolia shrub <1 microphyll nanophanaerophyte native Seringia nephrospermum shrub <1 microphyll nanophanaerophyte native Setaria verticillata tussock grass microphyll therophyte native Sida arenicola shrub <1 leptophyll nanophanaerophyte native Sida cardiophylla shrub <1 leptophyll nanophanaerophyte native Sida fibulifera shrub <1 leptophyll hemicryptophyte native
Sida platycalx shrub <1 leptophyll nanophanaerophyte native Sida sp. Golden calyces glabrous shrub <1 leptophyll nanophanaerophyte native Sida sp. spiciform panicles shrub <1 leptophyll nanophanaerophyte native Solanium horridum herb microphyll nanophanaerophyte native Solanum centrale shrub <1 microphyll nanophanaerophyte native Solanum ferocissimum herb microphyll nanophanaerophyte native Solanum lasiophyllum shrub <1 microphyll nanophanaerophyte native Solanum nigrum shrub <1 microphyll nanophanaerophyte weed
Sporoboluss caroli tussock grass microphyll therophyte native Stackhousia muricata subsp. annual herb leptophyll therophyte native
Stackhousia muricata subsp. Annual
herb leptophyll therophyte native
Streptoglossa bubakii herb leptophyll therophyte native Stylobasium spathulatum shrub <1 microphyll nanophanaerophyte native Themeda triandra tussock grass microphyll therophyte native Trachymeme oleracea herb microphyll therophyte native Triodia melvillei hummock grass microphyll therophyte native Triodia pungens hummock grass microphyll therophyte native Triodia vanleeuwenii hummock grass microphyll therophyte native Triodia wiseana hummock grass microphyll therophyte native Valchellia farnesiana shrub 1–2 nanophyll nanophanaerophyte weed Vigna sp Hamersley clay creeper nanophyll therophyte native
Table 3. Leaf-size classification (mm)
Leaf-size
Leptophyll
<0.25cm2 (<5mm x
5mm)
Nanophyll
>0.25<2.25 cm2
(between 5mm x
5mm and 1.5 x
1.5cm)
Microphyll
>2.25< 20.25cm2
(between 1.5cm x
1.5cm and 4.5 x
4.5cm)
Mesophyll
>20.25<182.25cm2
(between 4.5cm x
4.5cm and 13.5 x
13.5cm)
Macrophyll
>182.25<1640.25cm2
(between 13.5cm x
13.5cm and 40.5cm x 40.5cm)
Megaphyll
>1640.25 cm2
