HYDROLOGICAL PROCESSES

Hydrol. Process. 20, 37– 50 (2006)

Published online 1 August 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hyp.5897

Runoff and peak flow responses to timber harvest and

forest age in southern Chile
Andrés Iroumé,1 * Octavio Mayen2 and Anton Huber3
1 Universidad Austral de Chile, Faculty of Forest Sciences, Institute of Forest Management, PO Box 567, Valdivia, Chile
2 Universidad Austral de Chile, Faculty of Forest Sciences, School of Graduate Studies, PO Box 567, Valdivia, Chile
3 Universidad Austral de Chile, Faculty of Sciences, Institute of Geosciences, PO Box 567, Valdivia, Chile

Abstract:
Runoff and peak flows in three experimental catchments with different forest conditions were analysed in a rainy
temperate climate in southern Chile.
The hydrological effects of clearcutting a Pinus radiata plantation covering 79Ð4% of the La Reina catchment were
studied by analysing runoff and peak flows in the pre- and post-harvesting periods. Mean annual runoff increased
110% after timber harvesting. Clearcutting generated a 32% mean increase in peak flows.
Peak flow and runoff were examined in two adjacent catchments with different forest conditions. The older plantation
in Los Ulmos 1 increasingly consumed more water than the younger plantation established at Los Ulmos 2, whereas
differences in peak flows between these two catchments were not significant.
Runoff and peak flows comparisons not only reflected changes in forest cover, but also the effect of rainfall
characteristics during the study periods and the basins’ morphologies. Comparisons between pre- and post-harvesting
peak discharges must be undertaken with caution, because a previous analysis at La Reina using a partial set of data
overestimated changes in peak flows after timber harvesting. Copyright  2005 John Wiley & Sons, Ltd.

KEY WORDS runoff; peak flows; timber harvesting; plantation forestry; forest age; southern Chile

INTRODUCTION
Chile has 2Ð1 ð 10 ha of plantation forests and at least an additional 2 ð 106 ha of plantable land of which
6

500 000 ha are due to be planted by the end of this decade (CORMA, 2003). Although 90% of the existing
plantations were established on land that had been deforested for decades and afforestation had contributed
to reduce erosion processes (CORMA, 2003), silvicultural activities in these forests with exotic fast-growing
species generate important hydrological impacts. Gayoso and Iroumé (1995) indicated that most hydrological
impacts derive from site preparation, road building and harvesting, and Iroumé et al. (2005) showed that
the main hydrological effects arise from the afforestation of uncovered lands and large-scale clearcutting
operations that take place at the end of short-rotation growing cycles in environments characterized by intense
winter rainfalls and dry summers.
More than the 76% of Chilean forest production is exported (INFOR, 2003). To be competitive in
international markets, forest companies must certify their products. In 2003, about 350 000 and 1Ð2 ð 106 ha
were certified by Forest Stewardship Council and ISO 14 001 standards systems respectively (CORMA, 2003;
FSC, 2003). Through these standards, forest companies are committed to adopting best management practices
(BMPs) to reduce or mitigate their environmental impacts, so that the quantification and the monitoring of
BMP effects on water quantity and quality, erosion and sediment transport become relevant.
Previous research supports knowledge of hydrological effects associated with plantation forestry in Chile.
Interception losses from coniferous and broadleaved (including Eucalyptus spp.) forests were presented by

* Correspondence to: Professor Andrés Iroumé, Universidad Austral de Chile, Faculty of Forest Sciences, Institute of Forest Management,
PO Box 567, Valdivia, Chile. E-mail: airoume@uach.cl

Received 7 May 2004

Copyright  2005 John Wiley & Sons, Ltd. Accepted 15 December 2004
38 A. IROUMÉ, O. MAYEN AND A. HUBER

Huber and Iroumé (2001), while Huber et al. (1998) and Iroumé and Huber (2002) analysed interception
losses influenced by rainfall and forest characteristics (species, density, age). Huber and co-workers (Huber
and López, 1993; Huber et al., 1998) reported increased soil water consumption up to 3 m deep and
decreased percolation under plantation forests compared with grasslands. Nutrient fluxes in forest ecosystems
were presented in Oyarzún et al. (1998), Godoy et al. (2001, 2003), Uyttendaele and Iroumé (2002) and
Oyarzún and Huber (2003). Proposals of water conservation guidelines were presented by Gayoso et al.
(2000).
Little has been reported in Chile regarding runoff and peak flows, although results on increases in summer
runoff and peak flows after timber harvesting were presented in Iroumé et al. (2005).
Elsewhere, Troendle and King (1987), Stoneman and Schofield (1989), Ruprecht et al. (1991), Cornish
(1993), Ruprecht and Stoneman (1993), David et al. (1994), Bari et al. (1996), Burton (1997), Lane and
Mackay (2001) and Swank et al. (2001) showed that, after timber harvesting (and even after intense thinning),
annual streamflow increases significantly from pre-harvesting conditions. Changes in streamflow after timber
harvesting occurred when more than the 20% of the forest cover was reduced (Stednick, 1996). Annual runoff
increased by 10 to 120%, depending on the size of the clearcut area (Keppeler and Ziemer, 1990; Zimmerman,
1992; Fahey, 1994; Dye and Poulter, 1995; Swank et al., 2001). David et al. (1994) and Bari et al. (1996)
reported on the effects of timber removal only in the first years after final harvest, but Ruprecht and Stoneman
(1993) found them up to 12–15 years. Peak flows also increased after timber harvesting (Fahey, 1994; Jones
and Grant, 1996; Thomas and Megahan, 1998; Beschta et al., 2000; Caissie et al., 2002). Increases only for
small events have been reported by Whitehead and Robinson (1993), Ziemer (1998) and Caissie et al. (2002),
in a range between 14 and 48% by Harr et al. (1979), Fahey (1994) and Swank et al. (2001) while Jones
and Grant (1996) found 50 and 100% increase in peak flows for small and large catchments, respectively.
The effect of storm type is still controversial as Smith (1987) found that timber harvesting affects storms
of 100-year return periods whereas Whitehead and Robinson (1993) reported no significant differences on
flood peaks from forested and grassland catchments for large events. La Marche and Lettenmaier (2001) and
Beschta et al. (2000) reported that peak flow increased after timber harvesting for 5 and 10 year recurrence
interval events, whereas Thomas and Megahan (1998) did not detect any change for flows having return
intervals larger than 2 years.
Magnitude and duration of post-harvesting effects on base and peak flows depend on soil type, hillslope
steepness, aspect and lithology of the catchment, rainfall quantity, frequency and intensity, as well as on
the extension and type of forest operations and characteristics of the vegetation that re-establishes after
the harvesting. Flow increase was proportional to harvested area in a catchment (Hibbert, 1967), more
pronounced after clearcuttings than partial harvestings (Rothacher, 1970; Fahey, 1994), and more significant
in wet temperate regions (Keppeler, 1998). Major effects occur up to 3 years after logging. After stabilization
by vegetation regrowth, streamflow quickly returned to baseline levels (Fahey, 1994; Ruprecht and Stoneman,
1993; Keppeler, 1998).
In this study, the results of an analysis of runoff and peak flows in experimental catchments are presented.
Runoff and peak flows registered in pre- and post-harvesting periods and from catchments with different
forest cover are compared. The hypothesis tested is that the reduction of forest cover generates higher flows.
These results will add to the knowledge of the hydrological effects of plantation forestry in Chile and will
contribute to the adoption of management schemes compatible with environmental requirements.

MATERIALS AND METHODS

Study area
The study area comprises four experimental catchments located on the coastal mountain range in the Tenth
Region of Chile (Figure 1).

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 39

Figure 1. Location of Los Pinos, Los Ulmos 1 and 2, and La Reina catchments

The La Reina catchment (LR) is 34Ð4 ha large, faces north and is located 60 km west of the city of Osorno.
The two Los Ulmos catchments are also located on the coastal mountain range, but 45 km south of Valdivia.
Los Ulmos 1 (LU1) and Los Ulmos 2 (LU2) are 10Ð8 ha and 16Ð1 ha respectively and face southwest and
north respectively. Finally, Los Pinos is an 89Ð8 ha catchment located 22 km to the north of the city of
Valdivia. Other characteristics of these four catchments are presented on Table I.
The catchments are located in a rainy temperate climate with annual rainfalls ranging between 1566 and
3201 mm recorded during the study, with 73% of the annual total concentrated from May to October. Long-
term mean annual precipitation for the study area is represented by data registered in Osorno and Valdivia
weather monitoring stations, and is 1278 mm and 2293 mm respectively.

Study periods and forest covers

Runoff and peak flows were analysed at LR from 1997 to 2002. Between 1997 and most of 1999, 79Ð4%
of the catchment area was covered with a P. radiata (Monterey pine) plantation established in 1977, and
roads and riparian vegetation occupied 20Ð6%. This Monterey pine forest was clearcut between October 1999
and March 2000 and the area was planted with E. nitens in winter (June–July) 2000. For this study, LR was
considered under pre- and post-harvesting conditions for periods 1997–99 and 2000–02 respectively.
At the Los Ulmos site, runoff and peak flows were analysed between years 2000 and 2002. During these
years, 81% of the LU1 catchment area was covered with an E. nitens plantation established in 1997 and 19%
was in roads and riparian vegetation. Before planting, the site preparation techniques included fire to reduce
woody debris and residues of the forest previously clearcut.
During the first half of year 2000, the LU2 catchment was covered with the residues of a mature Monterey
pine plantation clearcut at the beginning of 1999. Woody debris and residues were mulched before E. nitens
(7Ð3 ha) and Monterey pine (3Ð7 ha) were planted in winter (June–July) 2000, covering 68% of the catchment
area. Roads, riparian vegetation and several stands of different species occupied the remaining 32% of the
catchment.

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
40 A. IROUMÉ, O. MAYEN AND A. HUBER

Table I. Characteristics of the studied catchments

Catchment Area Drainage Mean Altitude Road Soils Plantation description
(ha) density hillslope range density
(m ha
1 ) (%) (m a.s.l.) (m ha
1 ) Area Fraction Year of
(ha) of total establish-
area ment
(%)

Los Pinos 89Ð80 43Ð8 7Ð6 114–224 20 Derived from volcanic ashes 29Ð6 33a 1976–82
of intermediate to modern
age, deposited on meta-
morphic rocks. Considered
as a transition from loamy
to red clayed soils
LU1 10Ð80 165Ð1 12Ð0 175–230 139 Red clayed originating from 8Ð7 81b 1997
old volcanic ashes
deposited on the coastal
metamorphic complex
LU2 16Ð10 58Ð9 20Ð6 155–210 87 As in LU1 10Ð9 68c 2000
LR 34Ð35 78Ð8 23Ð7 35–225 12 Transition from those 27Ð3 79Ð4d 1977d
originating from old
volcanic ashes deposited
on volcanic conglomerates
and those derived from
old clays sedimented on
volcanic andesitic and
basaltic formations
2000

a Los Pinos: 33% adult Pinus radiata plantations established between 1976 and 1982, 40% grassland and 27% riparian vegetation. No forest
operations during the study period.
b LU1: Eucalyptus nitens plantation established in 1997 with 1600 trees/ha. The remaining 19% of the area corresponds to roads and riparian
vegetation, 19%.
c LU2: Eucalyptus nitens (7Ð3 ha) and P. radiata (3Ð7 ha) forest planted in winter (June–July) 2000. Roads, riparian vegetation and several
stands of different species account for the remaining 32% of the catchment area.
d LR: P. radiata plantation established in 1977 clearcut and replaced by a E. nitens forest planted in winter (June–July) 2000. Roads and
riparian vegetation correspond to 20.6%.

Los Pinos had no forest operations or land-use changes during the study and was considered as the
control.

Experimental equipments and data analysis

Precipitation in each catchment was registered on a half-hourly basis through digital tipping-bucket rain
gauges. In the Los Ulmos area, one rain gauge located halfway between the two catchments was used.
All the experimental catchments were controlled by purpose-built gauging stations equipped with pressure
transducers, recording water level on hourly basis (LR has a gauging station with flume, whereas all the others
have notch weirs).
The hydrological behaviour of the experimental catchments for the different study periods was compared by
analysing runoff and peak discharge. Only peak flows generated by individual storm events (the precipitation
time-series are divided into events based on the criterion that one event has ended and another event has
started if more than 5 h pass with zero rainfall) with total rainfalls higher than 5 mm were considered at both
the LR and Los Ulmos sites.

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 41

Differences  in runoff and peak flows from pre- and post-harvesting conditions at LR and between LU1
and LU2 were obtained using the following equation:

% D X2 /X1
1 ð 100 1

For the LR catchment, X1 and X2 represent data (runoff in millimetres and peak flows in litres per second)
from pre- and post-harvesting conditions respectively. At the Los Ulmos site, X1 and X2 refer to data (runoff in
millimetres or specific peak flows in litres per second per hectare) from LU1 and LU2 respectively. Differences
between X1 and X2 were analysed for statistical significance using a t-statistic (Freese, 1984) with a 95%
level of confidence.
At LR, runoff and peak flow were compared between periods with different rainfalls. Peak flow comparisons
were accompanied with t-statistic tests to determine statistical significance of the potential differences between
mean rainfall volumes of individual storms generating peak discharges during the pre- and post-harvesting
conditions. Runoff during 1997 and 2002 was tested for homogeneity using a double mass approach with the
Los Pinos as control.

RESULTS AND DISCUSSION

Precipitation
Total annual precipitation registered during the study period for the LR and Los Ulmos sites is presented
in Tables II and III respectively.
At the LR site, maximum and minimum annual rainfalls from 1997 to 2002 were recorded in 2002
(3511 mm) and 1998 (1566 mm); Table II. Years 1997 and 1998 were associated with El Niño–southern
oscillation meteorological perturbations, specifically the ‘El Niño’ and ‘La Niña’ phenomena (Dirección
Meteorológica de Chile, 2003).
At Los Ulmos, the highest and lowest annual rainfalls from 2000 to 2002 were registered in 2002 (3161 mm)
and 2001 (1623 mm); Table III.
Data from the Osorno and Valdivia weather stations indicate that years 1997, 2000, 2001 and 2002 were
wetter than the long-term average for the study area, whereas 1998 and 1999 were drier. Runoff and peak
flows data for the pre-harvesting condition at the LR site had one year with rainfalls above the average (1997)
and two below (1998 and 1999). All the hydrological information for post-harvesting conditions came from
above-average years. At the Los Ulmos site, runoff and peak flows data refer to three years (2000, 2001 and
2002) above the average precipitation.

Runoff at LR
Annual runoff ranged between 321 and 1653 mm during the pre-harvesting period and between 1773
and 2427 mm in the post-harvesting period (Table II). Mean annual runoff coefficients (annual runoff/annual
rainfall) were 40Ð1% during pre-harvesting conditions (range 20Ð5–51Ð6%) and 69Ð9% in the post-harvesting
period (range 69Ð1–71Ð6%).
The relationships between annual precipitation and annual total runoff for both pre- and post-harvesting
conditions showed that runoffs were higher after timber harvesting (Figure 2). Rainfall and runoff relationships
for the wettest (June) and driest (December) months followed similar patterns (Figure 3).
On average, annual runoff was 917 mm year
1 and 2033 mm year
1 during pre- and post-harvesting periods
respectively, resulting in a mean increase of 1116 mm year
1 (i.e. 122%) after the clearcut of the P. radiata
plantation that covered 79Ð4% of the catchment area.
The 122% increase in runoff may be partly due to the higher rainfall during the post-harvesting period
(on average, the annual rainfall was 621 mm year
1 or 27% higher than in the pre-harvesting period). The

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 42

Table II. Monthly precipitation P and runoff R between 1997 and 2002 at the LR catchment

1997 1998 1999 Mean 1997–99 2000 2001 2002 Mean 2000–02

Copyright  2005 John Wiley & Sons, Ltd.

P R P R P R P R P R P R P R P R
(mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm)
mm %a mm %a

January 185 90 34 16 25 4 81 37 45Ð1 29 11 209 38 30 9 90 20 21Ð7

February 84 65 24 8 80 5 63 26 40Ð7 212 45 57 21 117 13 129 26 20Ð3
March 13 1 75 12 63 6 50 6 12Ð6 101 28 131 36 187 40 140 34 24Ð6
April 393 77 148 31 49 2 197 37 18Ð6 118 31 91 20 234 71 148 41 27Ð6
May 171 55 218 33 187 22 192 37 19Ð0 146 67 484 223 445 256 358 182 50Ð8
June 595 333 262 32 550 147 469 170 36Ð3 822 531 555 539 381 326 586 465 79Ð4
July 705 449 202 56 219 73 376 193 51Ð3 414 358 605 573 324 210 448 380 85Ð0
August 302 219 303 96 491 261 365 192 52Ð5 171 233 236 227 511 403 306 288 94Ð0
September 191 147 148 27 225 167 188 114 60Ð5 215 175 104 111 253 231 191 172 90Ð2
October 290 106 39 3 45 47 125 52 41Ð9 137 141 46 46 622 571 269 253 94Ð1
November 169 67 69 1 72 25 103 31 30Ð2 127 103 110 36 286 203 174 114 65Ð4
A. IROUMÉ, O. MAYEN AND A. HUBER

December 102 44 43 5 86 21 77 23 30Ð1 66 51 22 30 120 93 69 58 83Ð5

Total 3201 1653 1566 321 2092 778 2286 917 40Ð1 2558 1773 2651 1898 3511 2427 2907 2033 69Ð9

a As a percentage of rainfall.

Hydrol. Process. 20, 37–50 (2006)

 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 43

Table III. Monthly P and runoff R at the Los Ulmos (LU1 and LU2) catchments from 2000 to 2001

2000 2001 2002

P LU1 LU2 P LU1 LU2 P LU1 LU2

(mm) (mm) (mm)
R (%) R (%) R (%) R (%) R (%) R (%)
(mm) (mm) (mm) (mm) (mm)

January 45 20 43Ð9 20 44Ð6 215 48 22Ð5 55 25Ð3 59 27 45Ð9 47 79Ð4

February 252 46 18Ð4 48 19Ð2 66 33 50Ð6 40 60Ð8 111 27 24Ð0 54 49Ð1
March 62 35 56Ð2 38 61Ð2 67 48 72Ð3 49 72Ð8 256 57 22Ð4 80 31Ð4
April 165 32 19Ð2 45 27Ð4 68 39 57Ð5 42 61Ð5 178 29 16Ð5 52 29Ð5
May 125 58 46Ð6 75 59Ð8 512 165 32Ð1 197 38Ð5 360 37 10Ð2 50 13Ð9
June 1070 585 54Ð7 574 53Ð7 513 357 69Ð5 339 66Ð0 321 54 16Ð7 79 24Ð5
July 419 312 74Ð5 318 75Ð8 520 352 67Ð8 326 62Ð8 306 342 111Ð6 327 106Ð7
August 166 261 157Ð5 244 147Ð1 318 279 87Ð6 276 86Ð8 372 337 90Ð6 355 95Ð4
September 237 158 66Ð7 176 74Ð4 118 122 102Ð8 129 109Ð0 258 239 92Ð4 242 93Ð9
October 159 108 68Ð1 132 83Ð4 84 102 121Ð1 109 130Ð1 508 255 50Ð2 250 49Ð2
November 136 111 81Ð2 121 88Ð9 148 47 31Ð4 70 47Ð2 285 163 57Ð1 237 83Ð2
December 72 75 103Ð7 68 94Ð2 15 31 215Ð3 41 284Ð6 149 125 84Ð1 134 90Ð3
Annual 2907 1800 61Ð9 1859 64Ð0 2645 1623 61Ð4 1673 63Ð3 3161 1690 53Ð5 1907 60Ð3

3000
Pre-harvesting
Post-harvesting

2000
Runoff (mm year-1)

1000

0
0 1000 2000 3000 4000
Precipitation (mm year-1)
Figure 2. Annual runoff at La Reina catchment: pre-harvesting (years 1997, 1998 and 1999) and post-harvesting (years 2000, 2001 and
2002) conditions

actual importance of timber harvesting is not easy to determine, although a reduction in interception and
transpiration rates certainly occurred after logging.
Two years with relatively similar annual rainfalls, one during the pre-harvesting period (1997, 3201 mm) and
the other in the post-harvesting condition (2002, 3511 mm), generated annual runoffs of 1653 and 2427 mm
respectively. Therefore, the increase in runoff (774 mm) was much higher than the differences in rainfall
(310 mm).

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
44 A. IROUMÉ, O. MAYEN AND A. HUBER

(a) 600 (b) 100

Pre-harvesting Pre-harvesting
Post-harvesting Post-harvesting
80
Runoff (mm month-1)

Runoff (mm month-1)

400
60

40
200

20

0 0
0 200 400 600 800 1000 0 20 40 60 80 100 120 140
Precipitation (mm month-1) Precipitation (mm month-1)
Figure 3. Runoff at La Reina catchment for wettest (June, a) and driest (December, b) months, in pre-harvesting (years 1997, 1998 and
1999) and post-harvesting (years 2000, 2001 and 2002) conditions

Lower runoffs in the 1997–99 period were related to the higher water consumption capacity of the plantation
forest during the pre-harvesting years. The effect of forest removal on runoff was then analysed using a double
mass approach comparing data from the LR and Los Pinos catchments (Figure 4). The significant increase
in gradient (from 0Ð73 to 1Ð41) of the graph of the post-harvesting period compared with the pre-harvesting
period indicates that more water was discharged from the catchment when the vegetation cover was removed.
The increase in runoff commenced at the beginning of February 2000, which coincided with the final period
of harvesting operations initiated in October 1999 (Iroumé et al., 2005).
Projecting the 1997–99 cumulated runoff trend beyond January 2000, it is possible to estimate ‘virtual’
annual runoff for years 2000, 2001 and 2002 as 802 mm, 1088 mm and 1175 mm respectively. Comparing

10000
Cumulated monthly runoff (mm), La Reina

y = 1.4134 x - 2124.8
8000 r2 = 0.9860

6000
2002

4000
y = 0.7292 x + 292.5
2001

r2 = 0.9459
2000

2000
1999
1998
1997

0
0 2000 4000 6000 8000 10000
Cumulated monthly runoff (mm), Los Pinos

Figure 4. Monthly runoff double mass curve analysis between Los Pinos and La Reina catchments for the 1997– 2002 study period

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 45

these estimations with the measured annual runoffs for the same years (1773 mm, 1898 mm and 2427 mm),
the double mass analysis indicates a mean increase of 1013 mm year
1 (971 mm, 810 mm and 1258 mm in
years 2000, 2001 and 2002, respectively).
As mentioned previously, mean annual runoff for the pre-harvesting period was 917 mm year
1 . After
timber harvesting the runoff increased on average by 1116 mm year
1 , but from the double mass analysis
a smaller amount (1013 mm year
1 ) could be attributed to the effect of forest removal, with the remaining
103 mm year
1 probably caused by the higher rainfall occurred during the post-harvesting period. Therefore,
on average, a 110% increase in runoff during the post-harvesting period can be associated with clearcutting
the P. radiata plantation that covered 79Ð4% of the catchment.
For this location and level of annual rainfall between 2000 and 2002, interception losses of 460 mm year
1
and transpiration of 570 mm year
1 have been measured in a 20–22-year-old P. radiata plantation (Huber and
López, 1993; Huber and Iroumé, 2001). The 1013 mm year
1 mean increase in runoff after timber harvesting
derived from the double mass analysis seems consistent with the elimination of the interception capacity and
the reduction in transpiration potential of the remaining vegetation compared with the previous forest cover.
The 110% increase in annual runoff after timber harvesting at LR lies in the upper part of the range reported
in the Introduction and reflects the size of the harvested area (79Ð4% of the catchment). Keppeler (1998) also
confirmed that flow increase after clearcutting was more relevant in wet temperate regions, such as the one
where LR is located. Finally, it is noteworthy that at the LR catchment the annual runoff increase is still
important at the third year after timber harvesting.

Runoff at Los Ulmos

For the 3 years of the period studied, annual runoffs in LU1 (forest planted in 1997, 81% of the area) were
lower than at the LU2 catchment (forest planted in 2000, 68% of the area); Table III. In 2000 and 2001 the
differences in runoff between LU1 and LU2 were 59 mm year
1 and 50 mm year
1 respectively, but in 2002
the difference was augmented to 217 mm year
1 , reflecting the higher water consumption of the Eucalyptus
plantation in its fifth year of growth.
In 2002, the plantation at LU1 should have intercepted between 320 and 470 mm year
1 and the vegetation
at LU2 no more than 100 to 150 mm year
1 (Huber and López, 1993; Huber and Iroumé, 2001). The difference
in interception losses between these two catchments (220–320 mm year
1 ) may explain the difference in
runoff. Runoff differences between LU1 and LU2 should reach a maximum in years 2004 or 2005, and then
decrease as the plantation in LU2 develops and becomes similar (in terms of interception and transpiration
capacities) to that in LU1.
During the driest months of summer (January to March), differences in runoff between LU2 and LU1
increased from 5Ð5 mm in 2000 to 70Ð7 mm in 2002 (Table IV), reflecting the increasing water consumption
of the plantation established in LU1 with higher interception and transpiration capacity (denser and with
deeper roots). In the wettest winter period (June to August) differences in runoff between LU2 and LU1 did
not show the same trend as in the driest months runoff in winter was higher in LU1 in 2000 and 2001 and
higher in LU2 in 2002. More intense and larger rainfall amounts, along with moister soils, probably reduce the

Table IV. Runoff in LU1, LU2 and their differences for the three driest (January to March) and wettest (June to August)
months
Period 2000 2001 2002

LU1 LU2  LU1 LU2  LU1 LU2 

(mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm) (mm)

January–March 100Ð7 106Ð2 5Ð5 129Ð9 143Ð0 13Ð1 110Ð8 181Ð5 70Ð7
June–August 1158Ð0 1135Ð3
22Ð7 987Ð8 941Ð4
46Ð4 732Ð1 759Ð8 28Ð7

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
46 A. IROUMÉ, O. MAYEN AND A. HUBER

differences in interception and transpiration capacity of the vegetation between the two catchments. Rainfall
characteristics in these winter months and the combined effect of catchment area (larger in LU2 compared
with LU1), drainage density (denser in LU1 than LU2), mean hillslope steepness (steeper in LU2 than in
LU1) and site preparation techniques (use of fire in LU1 and mulching in LU2) likely affected the differences
between runoff registered during the wetter months.

Combined analysis of runoff between LR and Los Ulmos

From 2000 to 2002 the LR and LU2 catchments featured a relatively similar vegetation cover. During these
years, runoff coefficients ranged between 69Ð1 and 71Ð6% in LR and 60Ð3 and 64% in LU2. A larger size (LR
being 34Ð4 ha and LU2 being 16Ð1 ha), higher percentage of clearcut area (79Ð4% in LR and 68% in LU2)
and steeper terrain (mean slope of 23Ð7% in LR against 20Ð6% in LU2) may explain the differences in annual
runoff coefficients.
Data from LR, LU1 and LU2 catchments were used to generate a relationship between annual runoff
coefficient and the degree of development of the plantation (Table V). The relationship (r 2 D 0Ð73) between
annual runoff (as a percentage of annual rainfall) and the number of years after the establishment of the
plantation shows a decreasing trend from about 69% the year after timber harvesting to 35% after 22 years
of plantation growth (Figure 5). The data used for the analysis come from years with different annual total
rainfall, but they clearly show a decrease in the annual runoff coefficient as the plantations increase their
water consumption capacities (i.e. interception and transpiration rates).

Peak flows at LR
Figure 6 illustrates the relationship between the size of rainfall events and the resultant peak flows at LR.
This relationship was considered independently for the pre- and post-harvesting periods, and in these cases
the r 2 values were 0Ð77 and 0Ð46 respectively, indicating moderate and lower correlation respectively.
The value of r 2 is lower for the pre-harvesting condition and reflects the higher variance between the size
of rainfall event and peak flows compared with the situation in the post-harvesting period (see Figure 6).
Peak flow-generation processes are very much affected by antecedent moisture conditions and rain total,
intensity and duration of each storm event. The type of vegetation influences rainfall interception and soil

100

y = - 1.4484 x + 66.919
r2 = 0.7316
Annual runoff (% of annual rainfall)

80

60

40

20

0
0 5 10 15 20 25
Years after plantation establishment
Figure 5. Annual runoff depletion associated with plantation growth

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 47

Table V. Annual runoff and years after the establishment of the

plantation

Annual runoff as Years after the Year and

fraction of establishment of catchment
annual rainfall (%) the plantation

64 0 2000, LU2
69 0 2000, LR
63 1 2001, LU2
72 1 2001, LR
60 2 2002, LU2
69 2 2002, LR
62 3 2000, LU1
61 4 2001, LU1
53 5 2002, LU1
52 20 1997, LR
20 21 1998, LR
37 22 1999, LR

400

r2 = 0.7707
300
Peak flow (L s-1)

r2 = 0.4620

200

100

0
0 100 200 300
Rainfall event (mm)
Pre-harvesting Post-harvesting

Figure 6. Peak flows at La Reina catchment

water retention. Deep-rooted trees with spreading branches may induce variation in interception and retention
that is not so pronounced as in the more homogenous cover that occurs after timber removal.
At the LR catchment, mean peak flows for the pre- and post-harvesting periods were 48Ð3 l s
1 and
63Ð8 l s
1 respectively (Figure 6). This difference (statistically significant as determined by the t-statistic at
the 95% level) represents a mean increase in peak flows of 32% after clearcutting the forest that covered the
79Ð4% of the area of this experimental catchment.
On average, the volume of precipitation from individual rainstorms that generated these peak flows during
the pre- and post-harvesting periods were not significantly different (t-statistic, 95% level), therefore supporting
the hypothesis that the increases in peak flows are associated with the differences in land cover between the
two periods.
The range of peak flow increases after timber harvesting were strongly correlated with the extension of
the clearcut area within a catchment, and the 32% increase found in this study is rather low considering the
harvested area at LR. Since afforestation effects in reducing peak flows are greater for smaller storms (Calder,
1992; Fahey, 1994), the precipitation characteristics in the study area (annual rainfall concentrated in winter

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
48 A. IROUMÉ, O. MAYEN AND A. HUBER

25
Specific peak flow (L s-1 ha-1)
20

15 r2 = 0.6422

10 r2 = 0.5973

0
0 100 200 300
Rainfall event (mm)
LU1 LU2

Figure 7. Peak flows at the Los Ulmos catchments

months and intense events) thus explain the lower impact of changes in forest cover on peak flows. In this
wet temperate region where the LR catchment is located, timber harvesting had a greater effect in increasing
annual runoff compared with peak flows.
Another important aspect to consider when providing information related to the increases of peak flows after
timber harvesting is the data sources. For the LR catchment, previous studies considered January 1998–August
1999 and March 2000–June 2001 as pre- and post-harvesting periods (Calisto, 2003) and then January
1998–August 1999 and March 2000–January 2002 as pre- and post-harvesting periods (Iroumé et al., 2002),
which reported increases in peak flows after clearcutting of 49% and 120% respectively. In these two analyses,
as well in the one reported here, the volumes of the individual rain storms showed no significant differences
(t-statistic, 95% level) between the pre- and post-harvesting periods, and the increases in peak flows after
timber removal were associated with the differences in land cover.

Peak flows at Los Ulmos and comparison with LR

Figure 7 reveals the relationship between the size of rainfall events and the resultant peak flows at the Los
Ulmos catchments. In each case the r 2 values were close to 0Ð6, indicating moderate correlation.
Mean specific peak flows were 2Ð6 l s
1 ha
1 and 2Ð7 l s
1 ha
1 at LU1 and LU2 respectively, representing
a non-significant difference (t-statistic, 95% level) of 2Ð1% (Figure 7).
Notwithstanding the differences in the area of forest cover between the two Los Ulmos catchments, the
more developed plantation in LU1 seems to have more effect on annual runoffs than in peak flows. Besides
the influence of forest cover, the differences in size (LU1 is smaller than LU2), hillslopes (LU2 is steeper than
LU1), drainage density (165 m ha
1 in LU1 and 59 m ha
1 in LU2) and forest road density (139 m ha
1 in
LU1 and 87 m ha
1 in LU2) between the two experimental catchments also affect peak flow processes.
The mean specific peak flow measured at LR during the post-harvesting period was 1Ð9 l s
1 ha
1 , lower
than those registered in LU1 and LU2 (2Ð6 l s
1 ha
1 and 2Ð7 l s
1 ha
1 respectively). This disparity may
possibly be explained in terms of size, as LR is much larger then the two Los Ulmos catchments.

CONCLUSIONS
During the 3 years following clearcutting of the P. radiata plantation that covered 79Ð4% of the LR catchment,
a 110% increase in annual runoff occurred, on average, and mean peak flows were 32% higher.

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
 RUNOFF RESPONSES TO TIMBER HARVEST AND FOREST AGE 49

In wet temperate regions with high annual rainfall totals concentrated during winter months, such as the
one where the LR catchment is located, timber harvesting has a greater effect in increasing annual runoff
than peak flows. Surface runoff is not observed in these catchments, and runoff processes are associated to
sub-superficial or groundwater flows, which are affected during forested conditions because of the higher
transpiration capacity of trees owing to their deeper root systems.
In the two Los Ulmos experimental catchments, the older plantation in LU1 increasingly consumed more
water than the younger (by 3 years) forest cover established at LU2 (59 mm in 2000 and 217 mm in 2002),
although the mean specific peak flows differences between these catchments were not significant.
Data from LR, LU1 and LU2 show a decrease in the annual runoff (as a percentage of annual precipitation)
as the plantations increase their water consumption capacities from about 69% the year after timber harvesting
to 35% after 22 years of plantation growth.
Reductions of forest cover appear to cause higher runoff and peak flows, although catchment morphology,
extent of data sets and rainfall characteristics occurring before and after timber harvesting should be carefully
considered to understand the hydrological effects of forest cover changes fully.

ACKNOWLEDGMENTS

This research was funded by Universidad Austral de Chile project DID S-200219.

REFERENCES
Bari MA, Smith N, Ruprecht JK, Boyd BW. 1996. Changes in streamflow components following logging and regeneration in the southern
forest of Western Australia. Hydrological Processes 10: 447–461.
Beschta RL, Pyles MR, Skaugset AE, Surfleet CG. 2000. Peakflow responses to forest practices in the western cascades of Oregon, USA.
Journal of Hydrology 233: 102–120.
Burton TA. 1997. Effects of basin-scale timber harvest on water yield and peak streamflow. Journal of the American Water Resources
Association 33: 1187– 1196.
Calder IR. 1992. The hydrological impact of land-use change with special reference to afforestation and deforestation. In Proceedings of
the Conference on Priorities for Water Resources Allocation and Management, Natural Resources and Engineers Advisers Conference,
Southampton, July; 91– 101.
Calisto JD. 2003. Estudio comparativo de caudales máximos en cuencas con diferentes cubiertas de bosque. Tesis Ing. For. Valdivia.
Universidad Austral de Chile. Facultad de Ciencias, Forestales.
Caissie D, Jolicoeur S, Bouchard M, Poncet E. 2002. Comparison of streamflow between pre and post timber harvesting in Catamaran Brook
(Canada). Journal of Hydrology 258: 232– 248.
CORMA. 2003. Chile forestal. http://www.corma.cl. [5 December 2003].
Cornish PM. 1993. The effects of logging and forest regeneration on water yields in a moist eucalypt forest in New South Wales, Australia.
Journal of Hydrology 150: 301–322.
David JS, Henriques MO, David TS, Tomé J, Ledger DC. 1994. Clearcutting effects on streamflow in coppiced Eucalyptus globulus stands
in Portugal. Journal of Hydrology 162: 143– 154.
Dirección Meteorológica de Chile. 2003. Los fenómenos climáticos ‘El Niño’ y ‘La Niña’. http://www.meteochile.cl [11 October 2003].
Dye PJ, Poulter AG. 1995. A field demonstration of the effect on streamflow of clearing invasive pine and wattle trees from a riparian zone.
South African Forestry Journal 173: 27–30.
Fahey B. 1994. The effect of plantation forestry on water yield in New Zealand. New Zealand Forestry 39(3): 18– 23.
Freese F. 1984. Statistics for Land Managers. Paeony Press: Scotland.
FSC. 2003. Forests certified by FSC-accredited certification bodies. http://www.fscoax.org [23 December, 2003].
Gayoso J, Iroumé A. 1995. Impacto del manejo de plantaciones sobre el ambiente fı́sico. Bosque 16(2): 34–56.
Gayoso J, Schlegel B, Acuña M. 2000. Guı́a de conservación de Agua. Facultad de Ciencias Forestales, Universidad Austral de Chile,
Valdivia, Chile.
Godoy R, Oyarzun C, Gerding V. 2001. Precipitation chemistry in deciduous and evergreen Nothofagus forest of southern Chile under a
low-deposition climate. Basic and Applied Ecology 2: 65– 72.
Godoy R, Paulino L, Oyarzun C. 2003. Atmospheric N deposition in central and southern Chile. An overview. Gayana Botanica 60(1):
47–53.
Harr R, Fredriksen RL, Rothacher J. 1979. Changes in streamflow following timber harvest in southwestern Oregon. USDA Forest Service,
General Technical Report PNW-249.
Huber A, Iroumé A. 2001. Variability of annual rainfall partitioning for different sites and forest covers in Chile. Journal of Hydrology 248:
78–92.

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)
50 A. IROUMÉ, O. MAYEN AND A. HUBER

Huber A, López D. 1993. Cambios en el balance hı́drico provocados por tala rasa de un rodal de Pinus radiata (D. Don), Valdivia, Chile.
Bosque 14(2): 11–18.
Huber A, Barriga P, Trecamán R. 1998. Efecto de la densidad de plantaciones de Eucalyptus nitens sobre el balance hı́drico en la zona de
Collipulli, IX Región (Chile). Bosque 19(1): 59–68.
Hibbert AR. 1967. Forest treatment effects on water yield. In Forest Hydrology, Sopper WE, Lull HW (eds). Pergamon Press: Oxford;
527–543.
INFOR. 2003. Estadı́sticas forestales 2002 . Instituto Forestal, Boletı́n Estadı́stico 88. Santiago, Chile.
Iroumé A, Huber A. 2002. Comparison of interception losses in a broadleaved native forest and a Pseudotsuga menziesii plantation in the
Andes Mountains of southern Chile. Hydrological Processes 16: 2347– 2361.
Iroumé A, Mayen O, Calisto JD. 2002. Efecto de la cubierta vegetal en los caudales máximos de cuencas de la décima región. Actas Primer
Congreso Chileno de Ciencias Forestales, Santiago, Chile, 23–25 October.
Iroumé A, Huber A, Schulz K. 2005. Summer flows in experimental catchments with different forest covers, Chile. Journal of Hydrology
300: 300– 313.
Jones JA, Grant GE. 1996. Peak flow response to clearcutting and roads in small and large basins, western Cascades, Oregon. Water
Resources Research 32(4): 959– 974.
Keppeler ET. 1998. The summer flow and water yield response to timber harvest . USDA Forest Service, General Technical Report PSW-168;
35–43.
Keppeler ET, Ziemer RR. 1990. Logging effects on streamflow: water yield and summer low flows at Caspar Creek in north-western
California. Water Resources Research 26(7): 1669– 1679.
La Marche JL, Lettenmaier DP. 2001. Effects of forest roads on flood flows in the Deschutes River, Washington. Earth Surface Processes
and Landforms 26: 115– 134.
Lane PNJ, Mackay SM. 2001. Streamflow response of mixed-species eucalypt forests to patch cutting and thinning treatments. Forest
Ecology and Management 143: 131– 142.
Oyarzún C, Huber A. 2003. Nitrogen export from forested and agricultural watersheds of southern Chile. Gayana Botanica 60(1): 63– 68.
Oyarzún C, Godoy R, Sepulveda A. 1998. Water and nutrient fluxes in a cool temperate rainforest at the Cordillera de la Costa in southern
Chile. Hydrological Processes 12: 1067– 1078.
Rothacher J. 1970. Increases in water yield following clear-cut logging in the Pacific Northwest. Water Resources Research 6: 653– 65.
Ruprecht JK, Schofield NJ, Crombie DS, Vertessy RA, Stoneman GL. 1991. Early hydrological response to intense forest thinning in
southwestern Australia. Journal of Hydrology 127: 261–277.
Ruprecht JK, Stoneman GL. 1993. Water yield issues in the jarrah forest of south-western Australia. Journal of Hydrology 150: 369– 391.
Smith PJT. 1987. Variation of water yield from catchments under grass and exotic forest, east Otago. Journal of Hydrology (New Zealand)
26: 175–184.
Stednick JD. 1996. Monitoring the effects of timber harvest on annual water yield. Journal of Hydrology 176: 79–95.
Stoneman GL, Schofield NJ. 1989. Silviculture for water production in jarrah forest of Western Australia: an evaluation. Forest Ecology and
Management 27: 273–293.
Swank WT, Vose JM, Elliot KJ. 2001. Long-term hydrologic and water quality responses following commercial clearcutting of mixed
hardwoods on a southern Appalachian catchment. Forest Ecology and Management 143(1): 163–178.
Thomas RB, Megahan WF. 1998. Peak flow response to clear-cutting and roads in small and large basins, western Cascades, Oregon: a
second opinion. Water Resources Research 34(12): 3393– 3403.
Troendle CA, King RM. 1987. The effects of partial and clearcutting on streamflow at Deadhorse Creek, Colorado. Journal of Hydrology
90: 145–157.
Uyttendaele GPY, Iroumé A. 2002. The solute budget of a forest catchment and solute fluxes within a Pinus radiata and a secondary native
forest site, southern Chile. Hydrological Processes 16: 2521– 2536.
Whitehead PG, Robinson M. 1993. Experimental basin studies—an international and historical perspective of forest impacts. Journal of
Hydrology 145: 217–230.
Ziemer RR. 1998. Flooding and stormflows. USDA Forest Service., General Technical Report PSW-168; 15–24.
Zimmermann RC. 1992. Impactos ambientales de las actividades forestales: orientación para su evaluación en los paı́ses en desarrollo. FAO
Guı́a para la Conservación No. 7. Organización de las Naciones Unidas para la Agricultura y la Alimentación. Roma.

Copyright  2005 John Wiley & Sons, Ltd. Hydrol. Process. 20, 37–50 (2006)