Skip to main content.

3.0 Results

3.1 Frequency of periphyton cover observations

Unwadeable conditions, high turbidity, and operational constraints prevented periphyton observations on an average of 25% of the monthly visits to the 73 individual sites included in the periphyton cover monitoring (range 3% to 76%, Appendix 2). The Waipaoa River (GS1) has particularly high turbidity (median 85 NTU) and periphyton observations were only possible on 24% of site visits. The Waitaki River was often highly turbid due to “glacial flour” and periphyton observations were only possible on 32% of visits at the SH1 site (TK6) and 51% of visits at Kurow (TK4). Other sites where periphyton observations were frequently impossible due to flow/clarity conditions were: Rangataiki at Te Teko (RO5, observations on 50% of visits), Mangakahia at Titoki (WH3, 39%), Wairua at Purua (WH4, 37%), Wanganui at Te Maire (TU1, 56%), Waipa at Whatawhata (HM2, 41%), Taieri at Tiroiti (DN1, 46%), Clutha at Balclutha (DN4, 44%), Mataura at Seaward Downs (DN5, 59%), Motu at Houpoto (GS4, 57%) and Wanganui at Paetawa (WA4, 56%) (Appendix 2).

Missing data, from site visits when flow conditions prevent periphyton assessment, probably inflated the mean recorded cover at sites above the actual mean because periphyton cover was likely to have been low when sites were not wadeable. However, annual maximum cover was unlikely to have been affected by missing data because high cover typically occurs during stable low flows when wading conditions are optimal.

3.2 Overall occurrence of obvious periphyton

The mean monthly riverbed cover of periphyton over 1990-2006 across all 73 sites was 4.9% for mats, 5.6% for filamentous growth, and 10.6% for total (filamentous + mat) cover (Fig. 3, Appendix 2), and median cover of both periphyton growth forms was zero i.e., on at least 50% of visits across all 73 sites there was no cover by either periphyton mats or filamentous growths. Mean total cover (filamentous + mats) only exceeded our proposed guideline (>40% cover) at the Ohinemuri (HM6) and upper Wanganui (TU1) sites. Although no sites had mean cover of filamentous algae or mats above the MfE guidelines (> 30% and > 60%, respectively, Biggs 2000), the maximum observed filamentous cover was exceeded at 78.1% of the sites (Appendix 2).

The mean annual maximum cover results give a more general picture of nuisance periphyton occurrence in these rivers: the mean for mats was 19.6% (median 18.0%, range 0 to 56.7%) and the mean for filamentous growths was 20.4% (median 14.6%, range 0.1 to 68.6%). The mean annual maximum cover for total periphyton cover (mats + filamentous cover) averaged 33.5% (median 32.6%, range 0 to 80.4%) (Fig. 3).

Figure 3: Total periphyton cover (filamentous + mats) of wadeable areas of riverbed at 73 NRWQN sites during 1990-2006 as (A) mean % cover and (B) mean annual maximum % cover

View maps | View text description
Figure 3: Total periphyton cover (filamentous + mats) of wadeable areas of riverbed at 73 NRWQN sites during 1990-2006 as (A) mean % cover and (B) mean annual maximum % cover

The mean annual maximum filamentous cover exceeded the MfE 30% guideline at 23.3% of the 73 sites, with more exceedence at impact (28% of impact sites), than baseline sites (6% of baseline sites) (Table 1, Appendix 2). The mean annual maximum total cover exceeded the recommended 40% threshold for aesthetic nuisance effects at 33% of sites (Appendix 2). Exceedence was more common at impact sites (45%) than baseline sites (13%). Over half of the sites had mat cover in excess of the MfE guideline on at least one occasion (i.e., maximum recorded >60%, Table 1, Appendix 2), but none had a mean annual maximum above the guideline level. Thus filamentous algae were the main cause of nuisance effects.

Table 1: Percentage of mean annual cover for the 73 sites monitored for periphyton between 1990 and 2006 that exceed the MfE guidelines (Biggs 2000)

Mean annual cover Guideline value used Percentage of sites by type exceeding the guidelines
Impact sites Pseudo-baseline sites Baseline sites All sites
Mean mat < 60 % 0 0 0 0
Mean filamentous < 30 % 0 0 0 0
Mean total < 40 % 2.6 12.5 0 2.6
Maximum mat < 60 % 0 0 0 0
Maximum filamentous < 30 % 28 25 6 23.3
Maximum total < 40 % 45 25 13 33

Annual maximum total periphyton cover at impact sites (mean = 40.2%) was significantly higher than at baseline sites (24.0%) (F 1,64 = 11.0, P = 0.001), as was annual maximum filamentous cover (means 25% and 13%, respectively, F 1,64 = 7.7, P = 0.007) and annual mean total cover (12.5% and 7.6% respectively, F 1,64 = 4.4, P = 0.04) (Table 2). Annual mean mat cover was also higher at impact (mean 5.8%) than baseline sites (mean 3.8%) (F 1,64 = 6.2, P = 0.016), but annual maximum mat cover did not differ significantly between impact and baseline sites (means and SD = 22.5% (14.6) and 16.6% (15.6), respectively).

Table 2: Comparison of mean percentage cover of periphyton between 42 impact and 23 baseline sites monitored between 1990 and 2006

Annual cover Mean per cent cover
Impact sites Baseline sites Significant difference
Mean mat 5.8 3.8 Yes (P = 0.016)
Mean filamentous 6.8 3.6 Yes (P = 0.001)
Mean total 12.5 7.6 Yes (P = 0.04)
Maximum mat 22.5 16.6 No (P = 0.059)
Maximum filamentous 24.6 13.0 Yes (P = 0.007)
Maximum total 40.2 24.0 Yes (P = 0.001)

3.3 Trends in obvious periphyton cover at NRWQN sites

Trend analysis for individual sites over the period 1990-2006 showed that the majority of the sites had statistically significant trends (at P < 0.1, without false discovery rate (FDR) adjustment for multiple comparisons) in annual mean or maximum periphyton cover as mat and/or filamentous (Appendix 3), with more sites having negative (decreasing) trends than positive (increasing) trends (Table 3, Fig. 4). For example, mean annual total periphyton cover decreased at 16 sites (22%) and increased at seven sites (9%) (Table 3).

The proportion of sites showing statistically significant increases (at P < 0.1, without FDR adjustment for multiple comparisons) in one or more classes of periphyton cover was similar among impact compared to baseline/pseudo-baseline sites (21% and 23%, respectively), whereas the proportion of sites showing declines was slightly larger amongst the impact sites (45%) compared to the baseline/pseudo-baseline sites (32%). Adjusting these trend analyses for a 5% FDR for more robust assessment of trends amongst the whole dataset (c.f. at individual sites) indicates significant declines in one or more classes of periphyton cover at five (16%) of the baseline sites and six (14%) of the impact sites and increases at two (5%) of impact sites (Table 3).

Figure 4: Trends in annual mean (left) and maximum (right) total periphyton cover (ie, mats + filamentous growths) 1990-2006 (as rs with year, P < 0.1, without false discovery rate adjustment).

View maps | View text description
Figure 4: Trends in annual mean (left) and maximum (right) total periphyton cover (ie, mats + filamentous growths) 1990-2006 (as rs with year, P < 0.1, without false discovery rate adjustment).

Note: See Table 3 and Appendix 3 for details on trends and site locations.

Table 3: Listing of sites where statistically significant trends in mean annual periphyton have occurred over the period 1990 to 2006

Filamentous mean cover Mats mean cover Total mean cover Type Site Filamentous maximum cover Mats maximum cover Total maximum cover
B AX1 ▲
I CH4 ▼
▼▼ I DN1 ▼
▼▼* ▼▼ ▼▼* B DN2 ▼▼*
▼ ▼ I DN4 ▼▼ ▼▼
▼ ▼▼ ▼▼ I DN5 ▼▼ ▼
▼▼ ▼▼* ▼▼ PB DN7 ▼▼ ▼▼* ▼▼
▼▼* ▼▼* ▼▼* I DN8 ▼▼ ▼▼* ▼▼
▼▼* ▼▼* ▼▼* B DN10 ▼▼* ▼▼ ▼▼*
▼ I GS1
▼▼* ▼▼ I GS4 ▼▼ ▼
▲ I GY1 ▲
▼▼ I GY2 ▼▼
▼▼* ▼ B GY3 ▼▼
▼▼* ▼▼ B GY4 ▼▼ ▼▼
▲▲ ▲▲ I HM2 ▲
▼▼ I HM4 ▼▼
▼▼ ▼▼ B HV1 ▼▼ ▼▼
▼ ▼ I HV2 ▼ ▼
▼ I HV3 ▼ ▼
▲▲ ▲ I NN1 ▲▲ ▲▲
▲▲ ▲▲ B NN2 ▲▲ ▲▲
B NN3 ▲
▼▼ ▲▲* ▲ I NN4 ▼ ▲▲ ▲
▲▲ ▲ PB NN5 ▲
▼▼ ▼ B RO1
▼ B RO4
▼ I RO5 ▲▲ ▲
▼▼* ▼▼ I TK1 ▼▼* ▼▼*
▼▼* I TK2 ▼▼*
▲▲ PB TK3
▼▼* ▼▼ I TK5 ▼▼ ▼▼
▼▼* I TK6 ▼▼
▲▲ ▲ I TU1 ▲▲*
B TU2 ▲▲
▼ I WA3 ▼
▲ ▲ I WA6 ▲▲ ▲▲
PB WA7 ▼
▼ B WH1 ▼
▼▼ I WH2 ▼▼
▼▼ ▼▼ I WH4 ▼ ▼
▼▼* I WN1 ▼▼* ▼▼
▲ B WN2 ▲
▼▼ I WN3 ▼
▲▲ I WN4 ▲

Note: See Appendix 2 for summaries of mean and annual maximum cover and Appendix 3 for a list of rs values for all sites. ▲ = increasing (0.05 < P < 0.1), ▲▲= increasing (P < 0.05), ▼ = decreasing (0.05 < P < 0.1), ▼▼ = decreasing (P < 0.05) for individual sites. * = significant correlations adjusted for a False Discovery Rate of 0.05 amongst correlations at all sites for each periphyton cover attribute. B = baseline, PB = pseudo-baseline; I = impact (Smith et al. 1989).

3.4 Downstream changes in periphyton cover between NRWQN sites

Twenty-four rivers had more than one site (usually an upstream “baseline” or “pseudo-baseline” and downstream “impact” site) on their length. These paired sites can be used to identify changes in periphyton conditions that might be linked to catchment activities between the two sites.

Table 4: Comparisons of total periphyton cover at paired sites and time trends (rs) in difference between 27 site pairs along 24 New Zealand river systems from 1990 to 2006

Paired sites (downstream – upstream) Mean of annual mean total cover Mean of the annual maximum total cover Time trends in up-down-stream difference in monthly total cover (rs) (fig. 5)
Hurunui CH2 – CH1# 8 – 5 40 – 35 0.28*
Waimakariri CH4 – CH3# 2 – 2 21 – 20 n.s.
Taieri DN3 – DN2# 15 – 12 57 – 36* 0.301*
Mataura DN5 – DN6# 18 – 2* 53 – 13* -0.41*
Oreti DN8 – DN7+ 11 – 1* 40 – 8* -0.34*
Waipaoa GS1 – GS2# 24 – 10 41 – 38 n.s.
Motu GS4 – GS3 16 – 13 53 – 56 -0.29*
Buller GY1 – NN5+ 6 – 4 24 – 24 -0.25
Grey GY2 – GY3# 11 – 7 35 – 27 n.s.
Waipa HM2 – HM1# 18 – 2* 41 – 10* 0.26
Waikato HM4 – HM3 11 – 10 36 – 33 -0.19
Ngarororo HV3 – HV4# 4 – 0* 22 – 3* n.s.
Mohaka HV5 – HV6+ 3 – 0 10 – 0 n.s.
Tukituki HV2 – HV1# 19 – 1* 62 – 5* -0.22
Motueka NN1 – NN2# 11 – 1* 60 – 10* n.s.
Wairau NN4 – NN3# 15 – 1* 64 – 8* n.s.
Rangitaiki RO5 – RO4# 8 – 17* 25 – 51* n.s.
Opihi TK1 – TK2 10 – 7 37 – 37 n.s.
Waitara WA1 – WA2# 17 – 1* 49 – 4* 0.19
Wanganui WA4 – TU1 5 – 43* 16 – 78* n.s.
Rangitikei WA6 – WA5+ 7 – 8 31 – 29 n.s.
Manawatu mid WA8 – WA7+ 7 – 16* 26 – 51* n.s.
Manawatu lower WA9 – WA8 14 – 7* 54 – 26* -0.32*
Hutt WN1 – WN2# 18 – 0* 66 – 1* -0.30*
Ruamahanga mid WN4 – WN5# 8 – 0* 37 – 4* n.s.
Ruamahanga lower WN3 – WN4 6 – 8 28 – 37 -0.235
Ruamahanga WN3 – WN5# 6 – 0* 28 – 4* -0.45*

Note: # = baseline; + = pseudo-baseline (otherwise impact, after Smith et al. 1989). Bold font indicates a statistically significant difference between each individual site pairs (paired-t test, P < 0.05). Statistically significant rs values of differences between site pairs over time since 1990 (P < 0.05) are listed. * = Statistically significant differences in means or correlations after adjusting for a False Discovery Rate of 0.05 amongst all sites for each periphyton cover attribute. n.s. = not statistically significant at P < 0.05 for time trends in difference between individual site pairs.

The majority of paired sites had more periphyton cover at the downstream site, although this did not always occur (Table 4). Thirteen of the 24 paired sites on individual rivers had significantly higher mean or annual maximum periphyton cover at the downstream site than at the upstream site (Table 4, paired t-tests, P < 0.05 with false discovery rate adjustment). The downstream increase in periphyton cover was particularly striking in the Tukituki, Motueka, Wairau, and Hutt Rivers. The cover at the upstream site in these rivers was low (1-10%) whereas the annual maximum observed total cover was typically equal to or greater than 60% at downstream sites (Appendix 2).

These downstream increases indicate that changes in catchment activities/characteristics between upstream (baseline) and downstream (impact) sites frequently result in increased periphyton cover, often to “nuisance” levels (greater than 40% total cover) at least once a year. It is likely that human activities between these sitess contribute to these increases. In some cases there are known point sources of nutrients (e.g., from municipal and industrial discharges to the Mataura and Manawatu Rivers), and intensive land-use practices are generally expected to cause downstream increases in diffuse nutrient inputs (Ballantine & Davies-Colley 2009; Larned et al 2004).

Only three rivers had lower total periphyton cover at the downstream sites than upstream (Table 4). High turbidity at the downstream Wanganui site (WA4) probably restricted the periphyton development by reducing light penetration to the riverbed (Davies-Colley & Nagels 2008). Along the Manawatu River, periphyton cover dropped between the site near Dannevirke (WA7) and downstream at Palmerston North (WA8), indicating changes in periphyton growth conditions, but increased again below wastewater discharges from Palmerston North and associated industries (WA9). The Rangitaiki River also had a higher cover upstream (RO4) than downstream (RO5), probably due to a higher streambed stability at the upstream site.

Streambed stability and turbid water probably also limited periphyton growth at several of the downstream sites that did not have higher cover than their upstream pair (e.g., Waikato at Rangiriri (HM4), Waipaoa at Kanakanaia (GS1)). However, the downstream sites on the Rangitikei, Grey, Hurunui, Waimakariri and Opihi Rivers had similar periphyton cover to their upstream paired site, indicating that changes associated with land management between these sites have not been large enough to cause increased periphyton growth.

The trends in the difference in percentage of total periphyton cover (mats + filamentous) between downstream and upstream sites on individual rivers provide another way of assessing trends in river health. Four paired-site comparisons had significant increasing trends and 10 paired-site comparisons had significant decreasing trends (Table 4, Fig. 5). Adjusting for a false discovery rate of 5% when viewing the data as a whole (c.f. as individual sites) revealed two significant increases in difference over time and six declines (Table 4). The increase in the differences through time for the Waipa, Taieri, Waitara, and Hurunui Rivers indicate that either the upstream sites have declining cover or the downstream sites have increasing cover. The individual site annual mean trends (Table 3) indicate that for the Taieri, the increase in difference was likely due to reduced cover at the upstream site (DN2), whereas in the Waipa the increased cover at the downstream site probably produced the increased difference between upstream and downstream sites over time.

Similarly, rivers with significantly decreasing differences between paired sites could result from the upstream having increasing cover or the downstream having decreasing cover. The trend towards smaller downstream increase in total cover along the Mataura appears to have been driven by reduced cover at the downstream site (DN5, see Table 3) and may be related to improved wastewater treatment and removal of significant point sources over the last decade. Significant trends of declining difference downstream on the Motu, Waikato, Hutt rivers appeared to be due to reductions in annual mean cover at the downstream sites on the (GS4, HM4 and WN1, Table 3).

Figure 5: Trends in difference in monthly total periphyton cover (filamentous + mats) between upstream and downstream sites at 24 NRWQN rivers as assessed for individual site pairs (without FDR adjustment)

View maps | View text description
Figure 5: Trends in difference in monthly total periphyton cover (filamentous + mats) between upstream and downstream sites at 24 NRWQN rivers as assessed for individual site pairs (without FDR adjustment)

Note: see Table 4 for details.

 

|