Channel esponse o mul iple damming in a meande ing i e , middle and lowe A agón 1
Ri e (Spain) 2
3
4
Ibisa e, A.1; Díaz, E.1; Olle o, A.2; Acín, V.3 & G anado, D.3
5
6
1 Dp . Geog aphy, P ehis o y and A chaeology, Uni e si y o he Basque Coun y, UPV/EHU 7
c/ Tomás y Valien e s/n, 01006 Vi o ia-Gas eiz, Spain 8
9
2 Dp . Geog aphy and Land Managemen , Uni e si y o Za agoza 10
c/ Ped o Ce buna s/n, 50009 Za agoza, Spain 11
12
3 ECOTER S.C. Ecology and Land, Za agoza, Spain 13
14
askoa.ib[email p o ec ed], Tl . +34945013996 Fax +34945013309 15
16
17
18
Abs ac 19
20
Small hyd opowe plan s (SHP) a ec i e low and sedimen anspo , and hus, impac i e 21
mo phology. Eigh hyd opowe schemes we e s udied along he meande ing middle and lowe eaches o 22
A agón Ri e (Spain) o assess hei e ec s on channel mo phology and sedimen dynamics om 1927 o 23
2010. GIS ools we e used o measu e changes in lu ial su aces, channel plan o m and la e al and 24
e ical dynamics. Th ee pe iods (ea ly, middle and la e 20 h cen u y) we e analysed o disce n he e ec s 25
o he main p essu es, such as changes in land use, la ge ese oi s ups eam, and SHP. Resul s we e 26
combined wi h ield and opog aphical measu emen s and hyd ological analysis. 27
28
Ac i e channel wid h and channel mig a ion su e ed a clea educ ion in he whole pe iod. They s a ed 29
as a consequence o land co e changes in he d ainage basin, bu hei speed inc eased a e a la ge 30
ese oi was buil ups eam. Mo e ecen changes occu ed since mos o he SHP we e pu in o 1
ope a ion in he 1990's, especially in hei sho -ci cui ed eaches and in he ou mo e downs eam ones. 2
These changes a e in e p e ed as a consequence o educed discha ge, ansi o y sedimen apping, and 3
eac i a ion o sedimen anspo a e wei s became illed as well as by he impac o lood hyd ology. 4
5
6
7
8
Keywo ds: channel adjus men ; human impac ; ese oi s; dams; small hyd opowe s a ion; lu ial 9
geomo phology; meande ing i e , A agón Ri e 10
11
12
13
In oduc ion 14
15
Wa e discha ge, especially high lows, and sedimen supply d i e lu ial dynamics and hus de e mine 16
i e channel o m (Simon, 1989; Downs, 1995; Ibisa e e al. 2011). The e o e, human ac i i ies a ec ing 17
discha ge o sedimen supply ha e impo an impac s in i e mo phology and adjus men s (Kondol , 18
1997; Rinaldi & Simon, 1998). These impac s include changes in land use (Ga cía-Ruiz e al., 1997 19
Liébaul & Piégay, 2002), ese oi s (Pe s, 1979; B and , 2000), wei s (Gu nell, 1997), g a el mining 20
(Pe i e al., 1996; Kondol , 1994), o mining was e (Knigh on, 1991). Addi ionally, o he in-channel 21
ac ions as channel modi ica ions (Simon & Rinaldi, 2000), o channeliza ions (B ookes, 1988) di ec ly 22
modi y channel o m and hus i e ecosys ems. 23
24
Small hyd opowe plan s (SHP, he e hose smalle han 50 MW) ypically di e pa o he wa e , use i 25
o p oduce elec ici y, and e e i o he i e downs eam. They can hus impac i e hyd omo phology 26
by a ec ing wa e discha ge and sedimen anspo , and lead o imbalances in i e mo phogene ic 27
p ocesses. The numbe o SHP has inc eased wo ldwide as a consequence o he need o enewable 28
ene gy sou ces (Alonso-T is án e al., 2011). Hyd opowe ac i i ies in lu ial sys ems modi y na u al 29
p ocesses, ea u es and o ms in i e s (Sea , 1995; Kondol & Ba alla, 2005; Csiki & Rhoads, 2010). 30
Modi ica ions caused by hyd opowe ac i i ies occu no only a he si e o he in as uc u es, bu also 1
ups eam and downs eam (Bake e al., 2011). 2
3
Al hough he e is a ai ly comp ehensi e in o ma ion on he mo phological impac s o hyd opowe 4
associa ed o la ge dams (Rolle , 2007), much less is known abou he impac s o wei s (Degou e, 2006; 5
Cou e , 1999) and SHP on hyd omo phological and ecological unc ioning (Wa ne , 2012; Mala oi, 6
2003, 2009). Some s udies explo e he ecological e ec s o small hyd opowe plan s on 7
mac oin e eb a e communi ies as a consequence o i e habi a agmen a ion (Fu e al., 2008; Co es e 8
al. 1998), o on ish (San os e al., 2006; Ha ield e al. 2007; Lewis e al., 2004). SHP seem o ha e a 9
smalle impac han ha o la ge ese oi s, bu because many i e s a e a ec ed by mul iple SHP, hei 10
cumula i e e ec could be impo an (Yang e al., 2011). The e o e, o i e manage s i is impo an o 11
de ine he hyd omo phological e ec s o hese in as uc u es and o pe o m es o a ion plans ha could 12
mi iga e o eco e lu ial dynamics. 13
14
In his s udy we analyse empo al changes in channel o m, in an a emp o dis inguish he mo phological 15
adjus men s caused by di e en ac o s ha a e go e ning lu ial p ocesses: land use, la ge ese oi s and 16
se e al SHP on a Spanish i e , he A agón. 17
18
S udy si e 19
20
The A agón Ri e is a 190 km-long ibu a y o Eb o Ri e ha d ains an 8609 km2 basin in he no h o 21
he Ibe ian Peninsula (Fig. 1). The Yesa ese oi , loca ed in he middle each o he A agón, has a s ong 22
e ec on wa e and sedimen discha ge downs eam (López-Mo eno, 2006; López-Mo eno e al., 2003). 23
The s udy a ea, loca ed downs eam om his ese oi , p esen s a g a el-bed meande ing channel. 24
Acco ding o Díaz & Olle o (2005), his 88 km-long sec ion can be di ided in o ou unc ional sec o s, 25
sensu Amo os & Pe s (1993), di e ing in alley se ing, plan o m and slope. The i s sec o , om Yesa 26
dam o he con luence wi h he I a i Ri e , has a slope o 0.3%, a sinuosi y index o 1.6, an a e age daily 27
discha ge o 44.9 m³/s, and a es ic ed loodplain exca a ed in i s own e aces. The second sec o 28
ex ends om he con luence wi h he I a i Ri e o Gallipienzo. I s mean daily discha ge is a ound 64 29
m³/s (Pe al a e al., 2003), he slope is 0.1% and he sinuosi y index 1.4, bu uns h ough a wide alley, 30
up o 1600 m, which allows la e al mobili y in he loodplain. The hi d sec o goes om Gallipienzo o 1
Mu illo, and he i e , wi h an es ima ed mean daily discha ge o 65 m³/s, uns h ough a na ow alley, 2
150-320 m wide, o e ia y conglome a es. The ou h and las sec o goes om Mu illo o he 3
con luence wi h he A ga Ri e , has an a e age annual daily discha ge o 68.6 m³/s, a slope o 0.1%, and 4
meande s eely h ough a loodplain wide han 3 km. 5
6
Fig. 1. S udy si e a ea and loca ion o he hyd opowe s a ions wi hin he A agón Ri e basin 7
8
The i e has been long a ec ed by impac s bo h on he d ainage basin and in he channel. Since he 9
beginning o he 20 h cen u y many de ences (g oynes, ip ap and bank s uc u es) ha e been cons uc ed 10
along he i e o educe channel mobili y and bank e osion and o p o ec om loods. These 11
in as uc u es encou aged human ac i i ies, mainly ag icul u e, in he loodplain. Addi ionally he i e 12
has been a ec ed by wo ese oi s, Yesa (446.9 hm³) buil in 1959 and I oiz (417 hm³) buil in 2004 in 13
he I a i Ri e , one o he main ibu a ies o he A agón (Fig. 1). The d ainage a ea o Yesa is a ound 14
25% o he o al o A agón basin while in I oiz is a 5.9%, and bo h ese oi s a e used mainly o 15
i iga ion pu poses. These ese oi s, bu especially Yesa, which is olde , ha e d as ically al e ed he 16
dynamics o he A agón Ri e , as a esul o educed discha ge (Fig. 2A) and sedimen apping (López-17
Mo eno; 2006, Acín e al., 2011). F om 1913 o 1956 he i e s ill p esen ed an almos na u al egime, 18
wi h a mean discha ge o 76.7 m³/s and highes discha ges du ing Ma ch. Mean daily discha ges 1
dec eased om 1959 o 2000 and especially du ing la e win e and sp ing, which also a ec ed o he 2
highes lows ha shi ed o Janua y. The annual mean discha ge was 64.2 m³/s. 3
Table 1. Main cha ac e is ics o s udied SHP (U: ups eam; S: sho -ci cui ed; D: downs eam) in he 4 middle and lowe eaches o A agón Ri e . Mean annual discha ges a e es ima ions based on he 5 discha ges ups eam and low da a o a i ing ibu a ies. (Da a sou ce: Eb o Wa e Agency and own 6 da a) 7
Hyd opowe
s a ion
Cons uc ion da e
Func ional
eaches
Mean
annual Q
(m³/s)
Maximum Q
di e ed in
canal (m³/s)
Wei
heigh
(m)
Reaches leng h (m)
U
S
D
Sangüesa I
1968 ( e o med in 1981)
1
44.9
14
2.5
1132
1948
536
Sangüesa II
1967 ( e o med in 1986)
2
64
16
3
825
132
1012
Cáseda
1927
2
64
8
2
1125
4000
950
Gallipienzo
1919 ( e o med in 1982)
3
65
40
5
1683
669
1051
Mu illo
1995
4
68.6
60
5
1408
2243
1718
San aca a
1994 ( e o med in 2005)
4
68.6
70
2
1146
4202
1447
Mélida
1995
4
68.6
77
2
489
3595
1715
Capa oso
2000
4
68.6
70
3
1069
3222
1128
8
Mo eo e , 8 small hyd opowe plan s (Fig. 1, Table 1) ha e been buil along 85 km o he middle and 9
lowe eaches o he A agón Ri e , which also abs ac wa e and modi y lowing discha ges in he sho -10
ci cui ed eaches. Some o hese plan s we e buil in he ea ly 20 h cen u y, bu we e e o med and made 11
mo e e ec i e a apping wa e in he 1980's and 1990's. The es we e buil in he las decade o he 20 h 12
cen u y. F om 2000 o 2010 he discha ge dec eased conside ably o a 42.9 m³/s o mean annual discha ge 13
o he o al amoun o bo h he di e ed and he sho -ci cui ed lows and a 22.7 m³/s when only he 14
sho -ci cui ed each was conside ed (Fig. 2A). Mo eo e lows p esen a e y high a iabili y linked o 15
hyd opowe equi emen s, which can be highe han 15 m³/s in hal an hou (Fig. 2B). Ne e heless, 16
unlike in he la ge ese oi s, mos o he wa e di e ed by he hyd opowe schemes e u ns downs eam 17
o he i e . 18
19
Fig. 2. Hyd ology o he A agón Ri e a Capa oso. A) Mean daily discha ge o h ee pe iods: 1913-20
1959 (be o e Yesa ese oi ), 1959-2000 (a e Yesa ese oi ) and 2000-2010 (a e Capa oso SHP). B) 21
Example o daily discha ge a ia ion du ing 24 hou s on 1s o Feb ua y o 2008 (Da a sou ce: Eb o 1
Wa e Agency). 2
3
Se en o hese SHP a e di e sion powe plan s. The meande s a e cu by means o a wei ha di e s pa 4
o he wa e in o a bypass channel, so called he di e sion channel. The wa e is e u ned downs eam in o 5
he na u al channel, so called he sho -ci cui ed channel (Fig. 3). Only in one case, Sangüesa II, he 6
hyd opowe plan is a un-o - i e powe plan wi h no di e sion channel. Thus, in mos hyd opowe 7
schemes h ee eaches can be di e en ia ed: ups eam om he wei ; he sho -ci cui ed each (S o ek e 8
al., 1996); and downs eam om he con luence wi h he e u ning bypass he i e ecei es again he ull 9
wa e discha ge. The leng h o he ups eam and downs eam eaches was de ined om he compa ison o 10
ae ial pho og aphies be o e and a e he cons uc ion o econs uc ion o he wei and ield 11
econnaissance in di e en low condi ions; ups eam was de ined by he dam e ec c ea ed by he wei 12
h ough he delimi a ion o he pool, he disappea ance o i les and he change in he wid h o he wa e 13
su ace, whe eas he downs eam each was delimi a ed by he end o he e ec o he apid o med by he 14
wa e eleased om he a i icial channel. The dimensions o he eaches hus a ied om one 15
hyd opowe s a ion o he o he , depending on wei size, he di e ed discha ge and he cha ac e is ics o 16
he i e channel and hyd opowe s a ion. The ups eam each anged om 825 m in Sangüesa II o 1683 17
m in Gallipienzo. The downs eam each leng h anged om 536 m in Sangüesa I o 1718 m in Mélida 18
(Table 1). 19
20
Fig. 3. Loca ion o he ou small hyd opowe plan s o he lowe sec o o he A agón Ri e . The h ee 21
eaches and hei in luence leng h a e ma ked. 22
23
The e ec o he ou hyd opowe s a ions se led in he lowe each o A agón Ri e is especially 24
ele an (Fig.3). All ou we e cons uc ed in he mid 1990’s (excep Mélida he es we e buil s a ing 25
om old wei s), ha e a simila discha ge concession, a ound 70 m³/s (Table 1) and a ec o y eigh 1
pe cen o he lowe A agón Ri e each aking in o accoun he dam e ec ups eam, he sho -ci cui ed 2
each, and he downs eam in luence. Fo his eason a deepe analysis has been pe o med on his lowe 3
each o he i e . 4
5
6
Da a and Me hods 7
8
We ollowed a s epwise p ocedu e o analyse he geomo phologic e ec s o hyd opowe plan s in he 9
A agón Ri e . Fi s , based on ae ial pho og aphs, we analysed he empo al changes in channel 10
dimensions, ipa ian su ace and plan o m in he eaches a ec ed by all 8 SHP. Then, we complemen ed 11
he s udy wi h ield su eys and hyd ological analyses. Finally a deepe s udy was pe o med in he 4 12
downs eam SHP. 13
14
The a ia ions in ac i e channel and ipa ian su ace we e s udied by compa ing ae ial pho og aphs and 15
o hopho o maps using GIS ools. Th ee ae ial ligh s o 1927, 1957, 1992 and an o hopho o map o 2010 16
we e used. O hose collec ed, 1927 is he i s ae ial pho og aphy a ailable a a scale app oxima ely o 17
1:9000; he ligh o 1957, a a scale o 1:33000, was aken jus a e he Yesa ese oi was buil ; he 18
1992 ligh , a a scale o 1:8000, was aken jus be o e he cons uc ion o mos o he hyd opowe 19
s a ions; and he o hopho o o 2010, a a scale o 1:5000, was he las a ailable. The ae ial pho og aphs 20
o 1957 we e al eady geo e e enced by he Go e nmen o Na a e. We co ec ed geome ically he ae ial 21
pho og aphs o 1927 and 1992 by means o he second-o de polynomial ans o ming echnique. In each 22
image and a e age o 12 pe manen g ound con ol poin s we e ma ched, spa ially well-dis ibu ed bu 23
wi h special concen a ion in he ea u es o in e es . The a e age oo mean squa e (RMS) e o o he 24
1927 pho og aphy was 8.3 m and 6.6 m o he 1992 one. All he images we e ec i ied using he cubic 25
con olu ion esampling (Chu ieco, 2002). 26
27
Fou basic lu ial ea u es lying wi hin he lu ial co ido we e iden i ied o measu e a ia ions in ac i e 28
channel and ipa ian su ace: wa e su ace, sedimen ba s (un ege a ed ba s which a e subme ged a high 29
discha ges), pionee ege a ion (ba s wi h pionee ege a ion o e aces wi h g ass), and ipa ian o es 30
(deposi ional o ms co e ed by ipa ian o es and inunda ed only du ing he la ges loods) (Su ian, 1
1999, Comi i e al., 2011). Bo h wa e su ace and un ege a ed sedimen ba s we e conside ed as he 2
ac i e channel, and he e o e aken as a single ea u e o analyse hei e olu ion. These ou 3
mo phological ea u es we e iden i ied h ough he isual analysis o ae ial pho og aphs and he la es 4
o hopho o map. The su aces o each mo phological ea u e and hei e olu ion we e delimi a ed using 5
GIS ools o he h ee eaches o each hyd opowe s a ion: ups eam, sho -ci cui ed and downs eam. In 6
he ou s a ions loca ed in he lowe each o A agón Ri e (Mu illo, San aca a, Mélida and Capa oso) 7
we calcula ed o he ou yea s, 1927, 1957, 1992 and 2010, he a e age ac i e channel wid h (Su ian, 8
1999; Rinaldi, 2003) o each 500-m each, and conside ed he wid h a bank ull s age (Pa ke e al., 9
2007; Zanoni e al., 2008). Addi ionally he e oded and cons uc ed su aces (Piégay e al., 2005) we e 10
calcula ed wi h GIS o h ee pe iods 1927-1957, 1957-1992 and 1992-2010. These mobilized su aces 11
we e measu ed conside ing he modi ied su ace and channel mig a ion be ween he ac i e channel in he 12
ae ial pho og aphy a he beginning o he pe iod and he posi ion o he ac i e channel a he inal o he 13
pe iod. Thus, we iden i ied he su aces gained (e osion a eas, sensu Ghoshal e al., 2010) and los 14
(deposi ion a eas) by he la e al shi o he i e . The esul s we e exp essed in a ios conside ing he 15
changes in su aces in ela ion o he leng h o each each o all SHP. 16
17
The compa ison o opog aphic maps o 1998 and 2008, a a scale o 1:5000, allowed he analysis o i e 18
bed ele a ion o he ou mo e downs eam SHP. Con ou lines loca ion was compa ed in bo h maps and 19
dis ances calcula ed. Longi udinal opog aphic p o iles we e done in he eaches ups eam om he wei 20
by means o a opog aphic le el (e o 0.02 m) o analyse he illing o he i e bed o sedimen s. 21
Besides, an exhaus i e ield econnaissance o he channel on bo h i e banks was ca ied ou , looking 22
o e idences o channel incision (exposed oo s, exposed ounda ions in b idges o wei s, i e bed 23
exca a ion, lineal and ups eam e osion symp oms) o agg ada ion (bu ied elemen s, bu ied base o 24
ipa ian ees) (Su ian, 1999; Rinaldi, 2003). 25
26
The hyd ological impac o hyd opowe s a ions was s udied based on he da a collec ed a he gauging 27
s a ion o Capa oso, which con inuously measu es discha ge a bo h he de i ed (bypass channel) and he 28
sho -ci cui ed each (na u al i e channel). The discha ge da a s a ed in 1912 and he de i ed discha ge 29
s a ed o be measu ed when he hyd opowe s a ion came in o ope a ion in 2000. Thus all he pe iod 30
analysed by he ae ial pho og aphs can be co e ed and in addi ion 10 yea s o bo h na u al i e (sho -1
ci cui ed) lows and di e sion da a a e a ailable o assess he hyd ological al e a ions de i ing om hese 2
in as uc u es. The impac o hyd opowe ac i i y was analysed o bo h mean annual discha ges and 3
lood hyd ology. The numbe o days in which he maximum di e sion discha ge is exceeded was 4
coun ed o he ou mo e downs eam SHP. Using Gumbel's law p obabili y dis ibu ion o he gauging 5
s a ion o Capa oso, we calcula ed he 5- and 10-yea e u n-pe iod lood discha ges (Q5 and Q10), and 6
Q2, conside ed he channel- o ming discha ge (Wolman & Mille , 1960; G ego y & Pa k, 1974). 7
8
C i ical discha ge o bedload en ainmen (Schokli sch, 1950) was calcula ed based on he g ain size o 9
he sho -ci cui ed each o each hyd opowe s a ion. The equency o days abo e he c i ical discha ge 10
o bedload en ainmen was analysed o he da a p o ided by he gauging s a ion o Capa oso. The 11
numbe o days ha his discha ge is exceeded was coun ed bo h in he sho -ci cui ed channel and he 12
o al discha ge ( he sum o he di e sion discha ge and he discha ge lowing h ough he sho -ci cui ed 13
each) o e alua e he impac o he di e sion. 14
15
16
Resul s 17
18
Changes in ac i e channel and ipa ian su aces and channel adjus men s 19
20
The A agón Ri e became na owe om 1927 o 2010: he su ace o he ac i e channel dec eased and 21
he su ace o ipa ian o es s inc eased, as can be app ecia ed in he examples o Mélida and Mu illo 22
hyd opowe s a ions (Fig. 4). This end was mo e e iden in he downs eam SHP (Mu illo, San aca a, 23
Mélida and Capa oso), whe e he a e age channel wid h dec eased in 79% (Fig. 5). This o e all end 24
can be spli in h ee pe iods (1927-1957, 1957-1992, and 1992-2010), acco ding o he main p essu es in 25
he basin. 26
1
F om 1957 o 1992 a ele an s abiliza ion and educ ion o la e al mo emen was app ecia ed in all 2
eaches (Fig. 8), wi h esidual e osion a eas con as ing wi h he wide deposi ion a eas. In a e age 9, 8 3
and 14 ha/km we e deposi ed in hei ups eam, sho -ci cui ed and downs eam eaches espec i ely, 4
while only 0.5, 0.3 and 0.5 ha/km we e e oded. Again, he eaches ini ially mos p one o mig a e 5
(San aca a, Cáseda, Capa oso and Mélida) su e ed he highes s abilisa ion a es. In all cases he 6
deposi ion o su aces domina ed. F om 1992 o 2010 deposi ion dec eased conside ably compa ed o 7
p e ious pe iods, as only 0.9, 1.4 and 2.3 ha/km we e a e age deposi ed in hei ups eam, sho -ci cui ed 8
and downs eam eaches espec i ely, while e osion inc eased wi h 0.9, 0.9 and 1 ha/km e oded 9
espec i ely. All he SHP cons uc ed in ha pe iod showed a d as ic educ ion o mig a ion a es 10
i espec i e o each ype (Fig. 8). 11
1
1
Fig. 8. E olu ion o e osion and deposi ion a es (ha/km), o he ups eam, sho -ci cui ed and 2
downs eam eaches in all SHP o he 1927-1957, 1957-1992 and 1992-2010 pe iods. E osion a ios a e 3
gi en in nega i e alues and deposi ion a ios in posi i e alues. 4
5
Compa ison o he 1998 and 2008 opog aphical maps showed ha in he ups eam eaches con ou lines 6
shi ed downs eam (2832 m in San aca a and 123 m in Mélida), hus showing agg ada ion o he i e 7
bed (Table 2). This p ocess was con i med by ield su eys (Fig. 9), al hough in his case he a e age 8
leng h o he agg aded each was 200 m. Mélida, he mos ecen SHP, was ela i ely less illed. All 9
sho -ci cui ed eaches p esen ed an ups eam shi o hei con ou line (703 m in Mu illo, 776 m in 1
San aca a, 963 m in Mélida and 265 m in Capa oso), showing an incision p ocess. Finally, in he 2
downs eam each he esponse was an agg ada ion o San aca a (123 m o shi downs eam) and 3
incision in Mélida (912 m o shi ups eam). The San aca a downs eam each ma ches wi h he ups eam 4
each o Mélida, being in luenced by he wei o Mélida, in p ocess o being illed. 5
6
Table 2. Shi in me e s o he con ou line wi hin each each om he compa ison o opog aphical 7 ca og aphy o 1998 and 2008 o Mu illo (MUR), San aca a (SAN), Mélida (MEL) and Capa oso 8 (CAP) SHP. The lack o da a means ha no con ou line c osses hose eaches. 9
MUR
SAN
MEL
CAP
Ups eam
+ 2832
+123
Sho -
ci cui ed
- 703
-776
-963
-265
Downs eam
+123
-912
10
11
12
Fig. 9. Longi udinal p o iles o he ups eam eaches. Dis ances measu ed om he wei in an ups eam 13
di ec ion (m). 14
15
Field e idences (exposu es in oo s, b idges and o he in as uc u es) showed an incision anging om 16
0.5 o 2 m in he sho -ci cui ed eaches (Table 3), p og essi e e osion om he wei and ups eam 17
e osion s a ing om he con luence o he di e ed wa e s. Only 8% o he e idences we e de ec ed 18
below he con luence, and no e idence o incision occu ed ups eam om he wei . The deepes incision, 19
deepe han 1.75 m, occu ed in Gallipienzo, he oldes SHP, whe eas he smalles incision occu ed in 20
Sangüesa II, he only un-o i e SHP. 21
Table 3. A e age incision in me e s in he sho -ci cui ed eaches a ec ed by hyd opowe s a ions (SAN 1 I: Sangüesa I; SAN II: Sangüesa II; CAS: Cáseda; GAL: Gallipienzo; MUR: Mu illo; SAN: San aca a; 2 MEL: Mélida; CAP: Capa oso). 3
SAN I
SAN II
CAS
GAL
MUR
SAN
MEL
CAP
1
0.7
1
1.75
0.9
1.3
0.9
1.1
4
5
Hyd ological changes 6
7
Mean daily discha ges dec eased in he h ee pe iods analysed (Fig. 2). A he sho -ci cui ed each in 8
Capa oso he gene al pa e n o he hyd ological egime was main ained, bu he a e age discha ge 9
dec eased in 45% ( om 40% in we yea s o 52% in d y ones). The pe cen age o wa e di e ed a ied 10
h oughou he yea om 34% in Oc obe o 62% in No embe (Fig. 10). Ri e discha ge exceeded he 11
wa e concession o SHP only in 39, 37, 31 and 27 days o Mu illo, San aca a, Mélida and Capa oso 12
espec i ely. 13
14
Fig. 10. Hyd ology in Capa oso. A) Sho -ci cui ed and di e ed mean annual discha ge (m³/s). B) Flood 15
e en in Ap il 2007. C) Peak low se ies om 1913 o 2011. D) Days abo e c i ical discha ge. Da a o 16
he sho -ci cui ed each a e compa ed o o al discha ge since 2000 when Capa oso SHP s a ed unning 17
(Da a sou ce: Eb o Wa e Agency). 18
1
Peak low se ies a Capa oso gauging s a ion (Fig. 10) showed a educ ion o he magni ude o loods 2
a e Yesa ese oi was buil . Q2, Q5 and Q10 discha ges also dec eased (Table 4), al hough hese esul s 3
p esen low signi icance due o b oad con idence in e als. Since Capa oso SHP s a ed ope a ing, 4
changes we e small, and limi ed o lowe magni ude e en s. Q2 discha ges dec eased in he h ee pe iods, 5
being his dec ease sligh ly mo e accen ua ed o he las pe iod. Howe e , Q5 and Q10 o o al discha ge 6
inc eased be ween 1957-1999 and 2000-2011. Meanwhile, in he sho -ci cui ed eaches Q5 dec eased 7
whe eas Q10 inc eased. The e o e, he compa ison be ween he discha ges lowing hese days h ough he 8
sho -ci cui ed sec ion and he o al discha ge, when he di e ed discha ge is added, showed scan 9
di e ences, which we e e en smalle as he discha ge inc eased. Peak low se ies also showed 10
mode a ely la ge loods in 2003, 2007 and 2010, wi h 928, 1027 and 797 m³/s espec i ely. 11
12
Table 4. Q2, Q5 and Q10 discha ges and con idence in e als o each pe iod a Capa oso gauging s a ion: 13 be o e ese oi s and SHP (1913-1956), since Yesa (1957-1999) and a e Capa oso SHP. 14
Q
2
(m³/s)
95 %
Con idence
in e al
Q
5
(m³/s)
95 %
Con idence
in e al
Q
10
(m³/s)
95 %
Con idence
in e al
1913-1956
592
± 83
1096
± 148
1275
± 203
1957-1999
549
± 55
718
± 99
836
± 136
2000-2011(sho -ci cui ed
discha ge)
434
± 161
698
± 314
877
± 441
2000-2011 ( o al discha ge)
483
± 160
752
± 312
930
± 439
15
16
The c i ical discha ge o bedload en ainmen (276.7 m³/s, and 1.05 y e u n pe iod) was exceeded only 17
a ew days pe yea , di e ences be ween he sho -ci cui ed each and he o al discha ge being mino 18
(Fig. 10). C i ical discha ge e olu ion h oughou all he da a se ies showed a educ ion o he numbe o 19
days since he mid 90’s, wi h a sligh di e ence be ween he o al and sho -ci cui ed discha ges. 20
21
22
Discussion 23
24
The main mo phological changes in he A agón Ri e consis ed on channel na owing, dec eased la e al 25
mo emen , and educed su ace o wa e and sedimen ba s, and an inc ease in he co e o ipa ian 26
o es s. Simila e olu ion has been desc ibed in o he i e s o he Eb o basin (Ma ínez-Cas o iejo e al. 27
1991; Begue ía e al. 2006; Olle o, 2010; Ma ín-Vide e al. 2010), wha sugges s a egional end owa ds 1
educed i e dynamism. Discha ge and lood in ensi y ha e s eadily dec eased in he Eb o basin du ing 2
he las decades oge he wi h an inc ease o he con ibu ion o low lows o he o al annual discha ge 3
(Begue ía e al. 2003; López-Mo eno e al. 2006), despi e no consis en end in he dis ibu ion o 4
p ecipi a ion (González Hidalgo e al., 2009; De Luis e al., 2009; Lo enzo, 2012). Reduced discha ge and 5
lood in ensi y has been hus a ibu ed o changes in land co e due o e o es a ion and abandonmen o 6
a mland (López-Mo eno e al. 2008), which also esul ed in dec eased sedimen supply (López-Mo eno 7
e al. 2003; Ga cía-Ruiz e al. 2010). Addi ional ac o s in luencing educed sedimen supply in he Eb o 8
basin include g a el mining (G anado e al., 2008; Ma ín-Vide e al. 2010), and ese oi s (López-9
Mo eno 2006). 10
11
Al hough i e dynamism dec eased in he A agón Ri e a leas since 1927, when he i s ae ial 12
pho og aphs we e aken, he a e o dec ease has a ied, and di e ences be ween pe iods can shed ligh 13
on he main causes unde lying his end. Be ween 1927 and 1957 he wid h o he ac i e channel 14
dec eased along he s udied a ea, whe eas he pionee ege a ion and ipa ian o es s inc eased. 15
Ne e heless, channel mig a ion was signi ican , wi h impo an a eas o e osion and deposi ion. As no 16
la ge ese oi s exis ed in he basin du ing his pe iod, he educed dynamism mus be a ibu ed mainly 17
o changes in land co e , especially inc ease in o es ed a eas. Fo es co e usually esul s in low 18
discha ge and dec eased sedimen yields, and e o es a ion has been shown o educe he dynamism o 19
i e s elsewhe e (Liébaul & Piégay, 2002). Bank p o ec ion, which s a ed in 1924 (Jimeno, 1996) and 20
use o ipa ian a eas o ag icul u e could ha e played an addi ional ole. The only wo SHPs in ope a ion 21
du ing his ea ly phase we e Gallipienzo and Cáseda, bu hei impac on i e dynamism seem o ha e 22
been mino , as hei e olu ion was e y simila o he es o he eaches s udied. This lack o impac 23
sugges s ha sedimen anspo was li le a ec ed by hei dams, p obably because du ing his pe iod 24
la ge loods we e s ill equen and he sedimen yield was high enough o ill he low dams in a sho 25
pe iod o ime. 26
27
F om 1957 o 1992 he end owa ds educed i e dynamism accele a ed, wi h a loss o ac i e channel 28
a ea o 1.7 ha/km/y, when i had p e iously been o 1.1 ha/km/y. This accele a ion coincides wi h he 29
building o he Yesa ese oi , which s a ed ope a ing in 1959 and has an impoundmen uno index (IR) 30
(i.e., ese oi capaci y espec o mean annual uno ) o 0.3. Many pape s ha e desc ibed a educ ion o 1
i e dynamism as a consequence o he cons uc ion o la ge ese oi s (e.g., G a , 2006; Go don & 2
Meen emeye , 2006). Two a e he main e ec s o la ge ese oi s on dynamism: sedimen s a a ion and 3
modi ica ion o he lood egime (Kondol , 1997; Kondol & Ba alla, 2005). Du ing 1960-85 he Yesa 4
ese oi los a 4.4% o i s s o age capaci y as i accumula ed 20.67 km³ o sedimen s (0.82 km³ pe yea , 5
López-Mo eno, 2006). Ne e heless, i seems ha he main e ec o he Yesa ese oi in ou s udy si es 6
is due o a educ ion o daily lows and lood equency. This e ec o educed lood equency is 7
consis en wi h he dec ease in e osional a eas measu ed a e he building o he Yesa ese oi along ou 8
s udy eaches. I is wo h no ing ha since he middle o he 20 h cen u y he dec ease o i e dynamism 9
was slowe ups eam om Yesa ese oi . The e he ac i e channel was educed 0.12 ha/km/y be ween 10
1946 and 1977 and 0.15 ha/km/y om 1977 o 2003 (G anado & Olle o, 2008). 11
12
F om 1992 o 2010 i e dynamism con inued o dec ease along he s udy a ea, bu especially so in he 13
lowe A agón sec o , whe e new SHPs we e buil . The e, ac i e channel wid h educ ion inc eased 14
p og essi ely downs eam, om 19% in Mu illo o 59% in Capa oso. This inc ease seems no o be a 15
esul o a cumula i e e ec o SHPs on i e dynamism, as p og essi e ac i e channel educ ion a ec s 16
mainly sho -ci cui ed eaches, no so clea in downs eam ones. Vadnais e al. (2012) also ound ha in a 17
complex sys ems dams and ese oi s do no p esen ed a cumula i e e ec on lu ial mo phology. 18
Howe e , a cumula i e e ec was alida ed in Gállego Ri e due o dams and g a el ex ac ions (Ma ín 19
Vide e al. 2010). We can only specula e abou he ac o s esponsible o his p og essi e change, as we 20
a e no awa e o any g adual change in en i onmen al condi ions be ween Mu illo and Capa oso, apa 21
om inc easing dis ance om he Yesa ese oi . Downs eam eaches adjus men s seem o be ela ed 22
wi h he local condi ions o he alley cha ac e is ics, as mo e inse ed channels (Mu illo and Mélida) 23
show a mo e mode a e wid h educ ion. Con as ing wi h his end o educed i e dynamism, e osion 24
inc eased du ing his pe iod, p obably as a esul o se e al lood e en s ha we e egis e ed, especially in 25
2003, 2007 and 2010. Channel e osion usually is associa ed o la ge loods (Kochel, 1988, Tu owski e 26
al., 2012), so he occu ence o hese e en s migh be he esponsible o he inc eased e osion du ing his 27
pe iod. 28
29
Fieldwo k in all SHP and analysis o opog aphic maps in he ou lowe mos SHP o measu e changes in 1
i e bed ele a ion yielded gene ally consis en esul s, and showed agg ada ion o he i e bed in 2
ups eam eaches as a consequence o sedimen e en ion in he dams, and deg ada ion in sho -ci cui ed 3
eaches. Low dams ap sedimen s, and hus, can esul in downs eam sedimen s a a ion and channel 4
incision (Ga cía-Ruiz & Puigde áb egas, 1984; Kondol , 1997; Doyle e al., 2005). Ne e heless, in 5
absence o ac i e managemen hei e ec s a e sho -li ed: as he dam becomes illed sedimen anspo 6
is esumed downs eam (B and , 2000; Bake e al., 2011; Csiki & Rhoads, 2010) 7
8
In summa y, SHP in he middle and lowe A agón Ri e seem o accen ua e a long end o educe i e 9
di e si y and los o geomo phological alue o he ee meande ing i e plan o m, de i ed om changes 10
in land uses and om he building o he la ge Yesa ese oi . Al hough he ime o ope a ion o mos 11
SHP is s ill oo sho o disce n he magni ude o he changes, we expec sedimen s a a ion o be 12
empo a y as he dams ill up. The impac s o al e ed hyd ology will las o as long as he SHP a e in 13
ope a ion, bu i is no clea how long will i ake o he i e o adjus o he new condi ions. Fu he mo e, 14
changes con inue o occu in he basin, such as he building o he I oiz ese oi , and hus i is expec ed 15
ha he Ri e A agón will con inue o change as a esul o he highly dynamic landscape. 16
17
Acknowledgmen s 18
This s udy is based on he esul s ob ained in he p ojec i led “E ec os de las minicen ales 19
hid oeléc icas en los sis emas lu iales de los íos A agón y A ga ( amos medio y bajo): diagnós ico, 20
endencias y p opues as de ges ión” (E ec s o he small hyd opowe s a ions in he lu ial sys ems o 21
A agón and A ga Ri e s (middle and lowe eaches): diagnosis, endencies and managemen p oposals). 22
We wan o hanks o Ges ión Ambien al de Na a a, S.A. o he unding o his esea ch, o Josu Elso, o 23
Camino Jaso o hei coope a ion in he p ojec and all he esea che s ha collabo a ed in he 24
geomo phological s udy (Ja ie Fuen es, Lau a E. Gonzalo, Daniel Balla ín, Daniel Mo a, Ra ael 25
Sánchez, Iban Sánchez, Ho acio Ga cía, Se gio Domenech). Finally we would like o hank he Spanish 26
Minis y o Economy and Compe i i eness o i s suppo h ough he p ojec SCARCE (Consolide -27
Ingenio 2010 CSD2009-00065). The au ho s also acknowledge suppo he Consolida ed Resea ch G oup 28
o he Basque Go e nmen IT 288-07, he Resea ch and Fo ma ion Uni “UFI11/09. Cua e na io: 29
Cambios Ambien ales y Huella Humana” and he Resea ch G oup o he A agón Regional Go e nmen 30
"Clima, Agua, Cambio Global y Sis emas Na u ales". We also wan o hank Kelly Janes om he 1
Uni e si y o Cali o nia, Be keley o he e iew o he ex in English, A u o Elosegi o his use ul 2
commen s and he e iew o he English and He é Piégay o his ele an ad ices and commen s as well 3
as he anonymous e iewe s ha helped o an impo an imp o emen o he pape . 4
5
Re e ences 6
7 Acín, V., E. Díaz, D. G anado, A. Ibisa e, A. & A. Olle o, 2011. Cambios ecien es en el cauce y la 8
llanu a de inundación del á ea de con luencia A agón-A ga (Na a a). Geog aphicalia 59-60: 11-25. 9
Amo os, C. & G.E. Pe s, coo ds., 1993. Hyd osys èmes lu iaux. Masson, Pa is. 10
Bake , D.W., B.P. Bledsoe, B.P., C.M. Albano & N.L. Po , 2011. Downs eam e ec s o di e sion dams 11
on sedimen and hyd aulic condi ions o Rocky Moun ain s eams. Ri e Resea ch and Applica ions 27: 12
388-401. 13
Begue ía, S., J.I. López-Mo eno, A. Lo en e, M. Seege & J.M. Ga cía-Ruiz, 2003. Assessing he E ec 14
o Clima e Oscilla ions and Land-use Changes on S eam low in he Cen al Spanish Py enees. Ambio 15
32(4): 283-286. 16
Begue ía, S., J.I. López-Mo eno, A. Gómez-Villa , V. Rubio, N. Lana-Renaul & J.M. Ga cía-Ruiz. 2006. 17
Flu ial adjus men s o soil e osion and plan co e changes in he cen al Spanish Py enees. Geog a iska 18
Annale 88 A (3): 177-186. 19
B and , S.A., 2000. Classi ica ion o geomo phological e ec s downs eam o dams. Ca ena 40: 375-401. 20
B ookes, A., 1988. Channelized Ri e s: Pe spec i es o En i onmen al Managemen . John Wiley & 21
Sons L d. 342 p. 22
Chu ieco, E., 2002. Telede ección ambien al. La obse ación de la ie a desde el espacio, A iel. 586 p. 23
Comi i, F., M. Da Canal, N. Su ian, L. Mao, L. Picco, M.A. Lenzi, 2011. Channel adjus men s and 24
ege a ion co e dynamics in a la ge g a el bed i e o e he las 200 yea s. Geomo phology 125: 147-25
159. 26
Co es, R., M.T. Fe ei a, S.V. Oli ei a & F. Godinho, 1998. Con as ing impac o small dams on he 27
mac oin e eb a es o wo Ibe ian moun ain i e s. Hyd obiologia 389 (1-3): 51-61. 28
Cou e , B., P. Le o , J.L. Pei y, P. Belleudy, 1999. La ges ion des i iè es. T anspo solide e 29
a e issemen s. Guide mé hodologique. Les é udes des agences de l’eau nº 65. Agence de l’Eau Rhône-30
Médi e anée-Co se, Pa is. 31
