UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Some effects of variation in weather and soil water storage on canopy evapotranspiration and net photosynthesis… Price, David Thomas 1987

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1987_A1 P73.pdf [ 33.81MB ]
Metadata
JSON: 831-1.0097498.json
JSON-LD: 831-1.0097498-ld.json
RDF/XML (Pretty): 831-1.0097498-rdf.xml
RDF/JSON: 831-1.0097498-rdf.json
Turtle: 831-1.0097498-turtle.txt
N-Triples: 831-1.0097498-rdf-ntriples.txt
Original Record: 831-1.0097498-source.json
Full Text
831-1.0097498-fulltext.txt
Citation
831-1.0097498.ris

Full Text

SOME EFFECTS OF VARIATION IN WEATHER AND SOIL WATER STORAGE ON CANOPY EVAPOTRANSPIRATION AND NET PHOTOSYNTHESIS OF A YOUNG DOUGLAS-FIR STAND By DAVID THOMAS PRICE B.Sc. Univers i ty of Edinburgh, 1975 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY i n THE FACULTY OF GRADUATE STUDIES Department of Forest Sciences We accept th i s thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA October 14, 1987 © David Thomas P r i c e , 1987 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date DE-6(3/81) - i i -ABSTRACT Measurements of the energy balances and net photosynthesis rates of two low product iv i ty coniferous forest canopies (12 and 22 years o l d ) , were made successfu l ly during both wet and dry growing seasons, using a modified Bowen Ratio method. Canopy conductances (g c) were calculated from canopy evaporation rates (E) using the Penman-Monteith equation. A model was devel -oped to predict canopy growth and evaporation rates from basic s o i l and weather data , and compared with the measured data . The photosynthesis model was phys io log ica l l y based, derived from recent work of Farquhar and coworkers. The canopy conductance model used an empirical approach, based on simple re lat ionships with recorded environmental va r iab les , while canopy E was predicted from the Penman-Monteith equation. Findings were: (1) Daytime E and canopy net photosynthesis rates (F £ ) were general ly lower in the younger canopy. (2) In the old canopy, E was more strongly decoupled from net i r radiance (R n) and more dependent on the atmospheric vapour pressure d e f i c i t (D) in accordance with the predict ions of McNaughton and Jarv is (1983). (3) In the old canopy, F c was s i g n i f i c a n t l y reduced by low s o i l water potent ia l (Y s) wi th in the range of s o i l water storages at which meas-urements were made, while g c was less dependent on Y g . From consideration of changes in i n t e r c e l l u l a r C0 2 concentration, g c was not found normally l i m i t i n g to F c . (4) No simple re lat ionship was apparent between solar i rradiance (S) and F at the canopy l e v e l . However highest F and canopy water use e f f i -ciency ra t ios occurred on cloudy days with low a i r temperature and low D. (5) Night-t ime F £ measurements indicated that canopy respi rat ion rates are general ly very high and hence a i r temperature was a major factor l i m i t i n g overa l l forest p roduct i v i t y . (6) The computer model could predict g from four var iables (D, S , root-zone s o i l water storage, W and time since dawn, t) with reasonable 2 success (r » 0 .75) . However, on days when g c was low, due to high 0 , E was occasional ly s i g n i f i c a n t l y in e r ro r , because the Penman-Monteith equation is more sens i t i ve to g £ when the l a t t e r i s low. Best agreement between measured and modelled E occurred on cloudy days when D was low and g c consequently h igh. (7) Values for the maximum rates of carboxy lat ion, as l imi ted by f o l i a r carboxylase a c t i v i t y and electron transport ra te , were set at one th i rd of those reported by Farquhar and coworkers, in order to obtain best overal l agreement between measured and modelled data. This requirement indicated that poor n u t r i t i o n was also l i m i t i n g to stand product i v i t y . (8) Model predict ion of canopy net photosynthesis was not s a t i s -2 factory (r * 0 .50) , a t t r ibuted mainly to using too simple an approach to estimating i r radiance at the indiv idual leaf l e v e l , and par t l y to unexplained var ia t ion in the measurements of F c . In spi te of i t s l i m i t a t i o n s , the model was found to respond r e a l i s t i c a l l y to changes in weather and Y , suggesting the approach was v a l i d , and might be more successful with further development. - iv -TABLE OF CONTENTS Page ABSTRACT i i TABLE OF CONTENTS iv LIST OF TABLES v i i i LIST OF FIGURES ix LIST OF SYMBOLS x v i i ACKNOWLEDGEMENTS x x i i i INTRODUCTION 1 CHAPTER I Measurement of Energy Balance, Evaporation and Net Photosynthesis of a Douglas - f i r Canopy 5 I. Introduction 5 I I . S i te Descr ipt ion and Experimental Procedures 7 1. S i te Descript ion 8 2. Stand Nut r i t iona l Status 9 3 . Weather 10 4. So i l Water Storage and Potent ial 11 5. Energy Balance/Bowen Ratio Measurements of Canopy Net Photosynthesis and Evaporation 14 1. Measurement System 14 2 . System Control and Data Co l lec t ion 19 3 . System Checks and Ca l ib rat ion 20 6. Calculat ions 21 1. Equations used 21 2. Data Reduction 27 - V -I I I . Results and Discussion 29 1. Ef fects of So i l Water Potent ial and Vapour Pressure D e f i c i t on Evapotranspiration Rate and Canopy Conductance of the Old Stand 31 1. So i l Water Potent ial 31 2 . Vapour Pressure D e f i c i t 38 2. Comparison of Diurnal 'Decoupling' between the Atmospheric Vapour Pressure D e f i c i t and Evaporation from the Young and Old Stand Canopies 39 1. Coupling in the Old Stand on Clear Days 40 2. Coupling in the Young Stand 41 3 . E f fects of Cloud Cover on Coupling in the Old Stand 45 4. Ef fects of Seasonal Changes in So i l Water Storage on a 1 and fi1 50 3 . Physical Factors inf luencing Canopy Net Photo synthesis and Respirat ion in the Old Stand 54 1. Ef fects of A i r Temperature 55 2. Ef fect of ¥ on Canopy Net Photosynthesis 59 3 . Ef fects of Cloudiness on Canopy Net Photosynthesis 65 4. Seasonal changes in Canopy Photo-synthesis and Water Use E f f i c iency 65 IV. Conclusions 68 V. References 70 CHAPTER II Development and Evaluation of a Model of Stand Canopy Evaporation and Photosynthesis 77 I. Introduction 77 I I . Structure of the Model 82 1. Solar Irradiance Submodel 83 2. Canopy Level Quantum Flux Density Submodel 84 3 . Leaf - leve l Stomatal Conductance Submodel 85 4. Leaf - leve l Photosynthesis Submodel.. . 93 5 . Canopy Level Calculat ions 99 - v i -l l i . Results and Discussion 101 1. Evaluation of the Conductance Submodel 101 2. Estimation of Canopy Evaporation from Modelled Canopy Conductance 109 3 . Evaluation of the Canopy Net Photosynthesis Model 119 4. Estimation of Canopy Water Use E f f i c iency Ratio from Modelled Canopy Evaporation and Net Photosynthesis 133 5. C r i t ique 141 IV. Conclusions 144 V. References 146 CONCLUSIONS 153 APPENDICES 157 I Ca lcu lat ion of the Psychrometric Constant, Y 157 II Measurement of Leaf Area Indices 160 III Ca lcu lat ion of the Aerodynamic and Canopy Resistance Terms 165 IV F o l i a r Nutrient Analyses for Douglas - f i r Needles at the Dunsmuir Creek S i te 173 V L i s t i n g of Data Analysis Program 175 VI Tables of Half -Hourly Data for Selected Days 198 VII Summary Table of Dai ly Means and Totals for So i l Moisture and Meteorological Variables Measured at Dunsmuir Creek 1983-1984 236 VIII L i s t i n g of Canopy Conductance/Photosynthesis Model 241 - v i i -IX Estimation of Forest Transpirat ion and C0 2 Uptake using the Penman-Monteith Equation and a Physio-log ica l Photosynthesis Model. (Draft text of Pr ice and Black 1987, Poster presented at International Union of Geodesy and Geophysics Workshop on Areal Evapotranspirat ion, August 1987, Vancouver, B.C.) 262 X Ef fects of Sala l Understory Removal on Photosyn-t h e t i c Rate and Stomatal Conductance of Young Doug las - f i r Trees. (Text of Pr ice e t a l . , 1986, Can. J . For. Res. 16: 90-97.) 289 XI Ef fects of Sala l Removal and Regrowth on S i te Water Balance and Tree Water A v a i l a b i l i t y in Young Douglas-f i r Plantat ions on Eastern Vancouver Is land. (Report by D.T. Pr ice to MacMillan Bloedel Woodlands Services D i v i s i o n , A p r i l 1987) 298 - v i i i -LIST OF TABLES Page 1.1 Approximate aerodynamic and canopy conductances, and derived midday values of McNaughton-Jarvis p. at Dunsmuir Creek compared with typ ica l values for a se lect ion of vegetation canopies in the l i t e r a t u r e . Conductance units are mm s _ 1 . Values of P. were calculated assuming T of 20 °C 53 1.2 Water use e f f i c i e n c y (WUE) rat ios at the Dunsmuir Creek s i t e in 1983 and 1984 calculated from F c/E and expressed on a mass percentage basis 57 2.1 Summary of 24 hour to ta l canopy evaporation, net photosynthesis and water use e f f i c i e n c y , measured and modelled, for the old stand on seven days at the Dunsmuir Creek s i t e 130 - i x -LIST OF FIGURES Page l . l a Canopy energy balance for the old stand on 20 August 1983. Plotted var iables are the so lar i r radiance (S) , net i r radiance ( R n ) » the sum of canopy heat storage and s o i l surface heat f lux dens i t ies (GQ + M) and the f lux dens i t ies of sensible heat (H), latent heat (LE) and photosynthetic f i x a t i o n energy (AF C ) . F i r s t of four selected c lear days showing e f fec t of decreasing W on canopy LE. So i l water content (W) was about 128 nm, p r o f i l e average to ta l s o i l water potent ia l (Y s ) was about -0 .035 MPa and maximum S was 840 W m" 2 at 12:15 32 1.1b Canopy energy balance for the old stand on 24 July 1984. W was about 84 mm, was about -0 .17 MPa and - 2 s maximum S was 876 W m at 12:15. [Notation as in F i g . l . l a ] 33 1.1c Canopy energy balance for the old stand on 29 July 1984. W was about 81 mm, Y 0 was about - 0 . 2 MPa and - 2 s maximum S was 882 W m at 12:15. [Notation as in F i g . l . l a ] 34 1.Id Canopy energy balance for the old stand on 20 August 1984. W was about 51 mm, Y„ was about s _? - 1 . 1 5 MPa and maximum S was 829 W m at 12:15. [Notation as in F i g . l . l a ] 35 1.2a Comparison of canopy evaporation rate (E) in res -ponse to atmospheric saturation vapour pressure d e f i c i t (D) for two ident ica l c lear days over the old stand, with very d i f fe rent W. During the 1983 growing season, W remained high and on 20 August was about 128 mm, whereas in 1984, W declined s tead i l y to about 51 mm by 20 August 36 - X -1.2b Comparison of canopy evaporation rate (E) in res -ponse to atmospheric saturation vapour pressure d e f i c i t (D) for two c lear days over the old stand during the 1984 growing season with e s s e n t i a l l y s i m i l a r W. On 24 July W was about 84 mm, decreasing to 81 mm on 29 J u l y , the small d i f ference being due to 5.6 mm p r e c i p i t a t i o n on 25 July 37 1.3a Canopy energy balance of the young stand on 9 August 1983, a c lear day. W was about 198 mm, Y was close to zero and maximum S was 855 W m~2 at 12:15. [Not-at ion as in F i g . 1.1a] 42 1.3b Canopy energy balance of the young stand on 11 August 1983, a cloudy day. W as about 196 mm, ¥ was about 0.0 MPa and maximum S was 824 W m~2 at 14:45. [Not-at ion as in F i g . 1.1a] 43 1.4 Comparison of canopy evaporation rate (E) in response to atmospheric saturat ion vapour pressure d e f i c i t (D) over the young stand in 1983 on 9 August (c lear) and 11 August (cloudy). Total canopy conductance, g c , was estimated from E, D and estimated canopy aero-dynamic res is tance, r A , using the Penman-Monteith combination equation 44 1.5a Canopy energy balance for the old stand on 26 August 1983. F i r s t of a pa i r of cloudy days showing e f fect of decreasing W on canopy LE. W was about 121 mm, Y was about -0 .045 MPa and maximum S was 825 W m~2 at 11:45. [Notation as in F i g . 1.1a] 45 1.5b Canopy energy balance for the old stand on 6 August 1984. Second of a pa i r of cloudy days showing e f fect of decreasing W on canopy LE. W was about 72 mm, ^ s was about - 0 . 3 2 MPa and maximum S was 702 W n f 2 at 12:15. [Notation as in F i g . 1.1a] 47 1.6 Comparison of canopy evaporation rate (E) in resp-onse to atmospheric saturat ion vapour pressure d e f i -c i t (D) for two cloudy days over the old stand, with very d i f f e r e n t W. On 26 August 1983 W was about 121 mm, whereas on 6 August 1984 W was about 72 mm 48 - x i -1.7 Seasonal changes in day-time a ' and midday fl1 r e l a -ted to W for the young stand in 1983 and for the old stand in la te August 1983 and 1984. For the purposes of t h i s study, a ' and Q' are defined as the P r i e s t l e y -Taylor a and McNaughton-Jarvis fl parameters appl ied to vegetated surfaces for varying root-zone s o i l water storage 51 1.8 Comparison of net canopy photosynthesis rates (F c ) and estimated canopy i n t e r c e l l u l a r C0 2 concentration (c^) in response to quantum f lux density (Q) and a i r temperature (T) for two c lear days over the old stand during the 1984 growing season, with e s s e n t i a l l y s i m i -l a r W. On 24 Ju ly W was about 84 mm, decreasing to 81 mm on 29 July and 51 mm on 20 August, the small change being due to 5.6 mm p r e c i p i t a t i o n on 25 Ju l y . The horizontal dotted l i ne plotted at -100 ug m~2 s - 1 on the graph of F £ indicates the probable contr ibut ion of s o i l respiratory C0 2 to F £ 56 1.9 Comparison of net canopy photosynthesis rates (F c ) and estimated canopy i n t e r c e l l u l a r C0 2 concentration ( c i ) in response to quantum f lux density (Q) and a i r temperature (T) for two ident ica l c lear days over the old stand, with very d i f fe ren t W. During the 1983 growing season, W remained high and on 20 August was about 128 mm, whereas in 1984, W decl ined stead-i l y to about 51 mm by 20 August. The horizontal dotted l i n e plotted at -100 ug m~2 s - 1 on the graph of F c indicates the probable contr ibut ion of s o i l respiratory C0 2 to F c 60 1.10 Comparison of net canopy photosynthesis rates (F c ) and estimated canopy I n t e r c e l l u l a r C0 2 concentration ( c . ) in response to quantum f lux density (Q) and a i r temperature (T) for two cloudy days over the old stand, with very d i f fe ren t W. On 23 August 1983 W was about 134 mm, whereas on 6 August 1984 W was about 72 mm. The horizontal dotted l i n e plotted at - 2 - 1 -100 ug m s on the graph of F c indicates the probable contr ibut ion of s o i l respiratory C0 2 to F £ 64 - x i i -1.11 Seasonal changes in d a i l y to ta l net canopy photo-synthesis ( F c ) , to ta l solar i r radiance (S) and mean d a i l y canopy water use e f f i c i e n c y ra t io (WUE). The dotted l i n e on the F„ p lot indicates the contr ibut ion c of s o i l resp i rat ion assuming a mean s o i l C0 2 e f f lux of 100 ug m~2 s " 1 . In the WUE p lot the s o l i d l i n e shows WUE calculated on a day-time basis ( ignoring night - t ime resp i rat ion and evaporation) while the dotted l i n e shows WUE calculated over 24 hours 66 2.1a Var iat ion of canopy conductance (g c ) with atmospher-i c vapour pressure d e f i c i t (D) for the old stand at Dunsmuir Creek during 1983 and 1984. A l l the p lotted points represent s ingle ha l f -hour l y values estimated from the measured evaporation rates using the Penman-Monteith equation. The s o l i d l i n e represents the "boundary l i n e " g c m o d = g c m a x 2.2434 exp[ -1 .6D 0 - 4 5 ] as used in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was smal 1 88 2.1b Var iat ion of canopy conductance (g c ) with s o i l water storage (W) for the old stand at Dunsmuir Creek dur-ing 1983 and 1984. A l l the plotted points represent s ingle ha l f -hour l y values. The s o l i d l i n e represents the "boundary l i n e " function g „ m r t . = g„ m a „0.3344W 0 - 2 3  J acmod acmax as used in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was smal 1 89 2.1c Var iat ion of canopy conductance (g c) with solar i r r a d -iance (S) for the old stand at Dunsmuir Creek during 1983 and 1984. A l l the plotted points represent s ingle ha l f -hour l y values. The s o l i d l i n e represents the "boundary l i n e " function g „ . = g / . m a v 2 . 0 4 S ~ 0 - 0 0 3 as J 3cmoa acmax used in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was smal l . Note that for the model, photon f lux density (Q) was used as a dr i v ing var iable even though the function shown here re lates g c to S. The constant 2.04 represents the average photon content of above-canopy so lar i r radiance in umol W - 1 90 - x i i i -2.2a Comparison of measured and modelled canopy conduct-ance for the old stand at Dunsmuir Creek on 20 August 1983 when s o i l water storage (W) was about 128 mm. The s o l i d l i n e given by Model A i s derived as the minimum of four conductance funct ions , whereas the dashed l ine given by Model B, estimates g as the product of these same four functions 102 2.2b Comparison of measured and modelled canopy conduct-ance for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for F i g . 2.2a] 103 2.2c Comparison of measured and modelled canopy conduct-ance for the old stand on 24 July 1984 when W was about 84 mm. [Notation as for F i g . 2.2a] 104 2.2d Comparison of measured and modelled canopy conduct-ance for the old stand on 29 July 1984 when W was about 81 mm. [Notation as for F i g . 2.2a] 105 2.2e Comparison of measured and modelled canopy conduct-ance for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2.2a] 106 2.2f Comparison of measured and modelled canopy conduct-ance for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2.2a] 107 2.2g Comparison of measured and modelled canopy conduct-ance for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2.2a] 108 2.3a Comparison of measured and modelled canopy evaporat-ion (E) for the old stand at Dunsmuir Creek on 20 August 1983 when s o i l water storage (W) was about 128 mm 110 2.3b Comparison of measured and modelled evaporation (E) fo r the old stand on 26 August 1983 when W was about 119 mm I l l 2.3c Comparison of measured and modelled evaporation (E) for the old stand on 24 July 1984 when W was about 84 mm 112 - x i v -2.3d Comparison of measured and modelled evaporation (E) for the old stand on 29 July 1984 when W was about 81 mm 113 2.3e Comparison of measured and modelled evaporation (E) for the old stand on 4 August 1984 when W was about 71 mm 114 2.3f Comparison of measured and modelled evaporation (E) for the old stand on 6 August 1984 when W was about 72 mm 115 2.3g Comparison of measured and modelled evaporation (E) for the old stand on 20 August 1984 when W was about 51 mm 116 2.4 Canopy evaporation rate as a function of canopy con-ductance, showing the sensit iv i ty to net irradiance, vapour pressure def ic i t and aerodynamic resistance. Conditions for a l l curves are: R„: 400 W rrf 2 and r. : _ i n A 10 s m , except where indicated, (a) D: 0.5 kPa; (b) D: 1.0 kPa; (c) D: 2.0 kPa; (d) same as (b) except R n : 300 W m - 1 ; (e) same as (b) except r A : 50 s rn"1. Curves (a) to (d) represent typical day-time values for the old stand while (e) shows some typical day-time values for the young stand 118 2.5a Comparison of measured and modelled canopy net photo-synthesis for the old stand at Dunsmuir Creek on 20 August 1983 when W was about 128 mm. Model A used values of v^m=. v a n d J m „ x , approximately one third of those cmd x max reported in the l iterature by Farquhar and coworkers, whereas Model B used values of approximately one f i f t h 121 2.5b Comparison of measured and modelled canopy net photo-synthesis for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for Fig. 2.5a] 122 2.5c Comparison of measured and modelled canopy net photo-synthesis for the old stand on 24 July 1984 when W was about 84 mm. [Notation as for F ig. 2.5a] 123 2.5d Comparison of measured and modelled canopy net photo-synthesis for the old stand on 29 July 1984 when W was about 81 mm. [Notation as for F ig. 2.5a] 124 - XV -2.5e Comparison of measured and modelled canopy net photo-synthesis for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2.5a] 125 2.5f Comparison of measured and modelled canopy net photo-synthesis for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2.5a] 126 2.5g Comparison of measured and modelled canopy net photo-synthesis for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2.5a] 127 2.6a S e n s i t i v i t y of canopy net photosynthesis model to changes in so lar i r rad iance , and stomatal conduct-ance, at 12:15 PST on 24 July 1984, a hot, c lear day. The s o l i d curves represent the C0 2 demand function for each canopy layer , where the uppermost l i n e represents the top layer . Lower l ines indicate the e f fec ts of reduced electron transport rates in lower canopy l a y -ers due to l i g h t absorption by upper layers . The dashed l ines represent the C0 2 supply functions cor re -sponding to measured ( r ight) and modelled ( l e f t ) mean layer g g . The intersect ion points enable the layer C0 2 f lux dens i t ies to be estimated (excluding "day-time dark r e s p i r a t i o n " , R d ) . Conditions were: S: 876 W r r f 2 , T: 30.0 °C, measured g c : 4.36 mm s " 1 , modelled g c : 3.36 mm s , modelled R d : 107 ug m~2 s " 1 per canopy layer , estimated s o i l respiratory C0 2 e f f l u x , R g : 100 ug m" 2 s - 1 131 2.6b S e n s i t i v i t y of canopy net photosynthesis model to changes in so lar i r rad iance , and stomatal conduct-ance, at 12:15 PST on 6 August 1984, a predominantly cool and cloudy day. Conditions were: S: 709 W rrf , T: 17.3 °C, measured g : 13.6 mm s " 1 , modelled g „ : - 1 c _2 _ 1 c 9.8 mm s , modelled R r f : 57 ug m s per canopy layer , estimated s o i l respiratory C0 2 e f f l u x , R g : 100 ug m~2 s " 1 . [For further explanation see also F i g . 2.6a] 132 - x v i -2.7a Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand at Dunsmuir Creek on 20 August 1983 when s o i l water storage (W) was about 128 mm. Night-t ime values were omitted because low atmospheric d i f f u s i v i t i e s and low evaporation rates combined with large resp-i ra to ry f luxes resulted in very large and e r r a t i c r a t i o s . Model A used values of V / . m a v and J m 3 „ approx-cmax max imately one t h i r d of those reported in the l i t e r a t u r e by Farquhar and coworkers, whereas Model B used values of approximately one f i f t h 134 2.7b Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for F i g . 2.7a] 135 2.7c Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 24 July 1984 when W was about 84 mm. [Notation as for F i g . 2.7a] 136 2.7d Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 29 July 1984 when W was about 81 mm. [Notation as for F i g . 2.7a] 137 2.7e Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2.7a] 138 2.7f Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2.7a] 139 2.7g Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2.7a] 140 - x v i i -LIST OF SYMBOLS Symbol D e s c r i p t i o n Roman symbols A U n i t s c i c c. c d D e E g c. g c ' g s ' 9 s ' water use e f f i c i e n c y , AFC/LE ambient atmospheric C0 2 concentration f o l i a r i n t e r c e l l u l a r C0 2 concentration s p e c i f i c heat of dry a i r AFC/H volumetric heat capacity of dry a i r volumetric heat capacity of top 50 mm s o i l zero plane displacement height ambient saturat ion vapour pressure d e f i c i t binary molecular d i f f u s i o n c o e f f i c i e n t s for C 0 2 , water vapour in dry a i r ambient vapour pressure saturat ion vapour pressure canopy evaporation rate canopy net C0 2 f lux density g rav i ta t iona l accelerat ion ( B 9.81) canopy (stomatal) d i f f u s i o n conductances for water vapour, C0 2 f o l i a r stomatal d i f fus ion conductances for water vapour, C0 2 s o i l heat f lux density at surface s o i l heat f lux density at 50 mm depth average tree height W W -1 -3 V9 m ug m" 3 J g K w w"1 J n f 3 K"1 -3 -1 kJ m K m kPa mm2 s~* kPa kPa _-2 -1 mg m_j s mm d m s -2 1 g m d mm s -1 mm s _ 1 umol m s~ W m" W m m -2 - x v i i i -H canopy sensible heat f lux density W m -2 H $ s o i l heat storage in top 50 mm of p r o f i l e J m J maximum potent ia l ( l i gh t saturated) rate of whole chain electron transport umol nf k von Karman's constant ( « 0.4) k Michaelis-Menten constant for ac t i va t ion of r ibulose diphosphate carboxylase (Rubisco), for constant pH and Mg cone. Pa k e f fec t i ve Michaelis-Menten constant for C0 2 (as substrate for Rubisco) Pa K. , K. canopy ext inct ion c o e f f i c i e n t for d i r e c t , d i f fuse shortwave radiat ion 2 - l K c , K y eddy d i f f u s i v i t i e s for C 0 2 , water vapour m s < H , <M eddy d i f f u s i v i t i e s for sensible heat, 2 . momentum m s 1 Monin-Obukhov s t a b i l i t y length m L latent heat of vaporisation of water kJ g - 1 L leaf area index of a s ingle canopy layer ? ? (plan area basis) m m 2 -2 L A canopy leaf area index (plan area basis) m m _? LE canopy latent heat f lux density W m _2 M canopy heat storage f lux density W m M , M , M molecular masses of dry a i r , C 0 2 , water . vapour g mol M H , M. rate of sens ib le , latent heat storage in ? canopy a i r W m M„. rate of sensible heat storage in canopy _ 2 biomass W m N neutron hydroprobe measurement count (30 s) N $ neutron hydroprobe standard count (30 s) p ambient p a r t i a l pressure of C0o Pa ca c P c i i n t e r c e l l u l a r p a r t i a l pressure of C0 2 Pa P_ ambient barometric pressure kPa - x i x -P . , P. quantum energy content of d i r e c t , d i f fuse . shortwave radiat ion with in canopy umol W" P gross photosynthesis rate (net of photo- _ 2 , 9 respi rat ion) with in chloroplast umol m" s P net photosynthesis rate with in 2 x chloroplast umol m s P g standard atmospheric pressure (= 101.3) kPa -2 - l Q photon f lux density (400-700 nm waveband) umol m s r. canopy aerodynamic resistance (assuming . s i m i l a r i t y for K u , K , K ) s m n v c r canopy d i f f u s i o n resistance (= g_ - 1 ) s rn"1 R c , R y gas constants for C 0 2 , water vapour J g _ 1 K - 1 -2 R d dark (or "day") respi rat ion rate ug m s R f proportion of canopy resp i rat ion coming ? from fo l iage yg m s~ -7 R growth resp i rat ion rate ug m s y R k canopy maintenance resp i rat ion c o e f f i c i e n t -? R assumed maximum canopy resp i rat ion rate ug m s max -? R m f canopy f o l i a r maintenance resp i rat ion rate ug m s R canopy stem and branch ( i . e . non - fo l ia r ) _ ? maintenance resp i rat ion rate ug m s~ _9 R n canopy net i r radiance W m -? R s o i l respiratory e f f l u x ug m, s , 5 g m~^  d ' 1 * 1 s , s de /dT evaluated at T, T kPa K w w _2 S measured so lar i r radiance at top of canopy W m S. d i rec t so lar i r radiance at top of canopy _ ? calculated from (S - S d ) W m S . . d i rec t so lar i r radiance at top of canopy ? estimated from p a r t i t i o n i n g of S t t W m S. d i f fuse so lar i r radiance at top of canopy ? calculated from S, S b t and S d t W m - XX -S . . d i f fuse so lar i r radiance at top of canopy _ 2 estimated from p a r t i t i o n i n g of S t t W nf S . . modelled so lar i r radiance at outer edge of _ 2 atmosphere, received p a r a l l e l to surface W m _ 2 Sp S at Dunsmuir Creek s i t e MJ m s - 2 - l S N S at Nanaimo AES weather s tat ion MJ m s t time s , d , y T ambient dry bulb temperature °C T. i ce -po in t temperature (= 273.16) K T K absolute temperature K T p e f reference temperature for estimating r* °C Tw ambient wet bulb temperature °C T $ mean s o i l temperature (0-50 mm depth) °C u mean horizontal wind ve loc i t y m s - 1 u* f r i c t i o n ve loc i t y m s - 1 V apparent maximum rate of carboxylat ion, 2 l im i ted by Rubisco a v a i l a b i l i t y umol nf s V cmax maximum potent ia l (RuP2 saturated) rate 2 of carboxy lat ion, umol m s V. apparent maximum rate of carboxylat ion, _ 2 J l imi ted by electron transport rate umol m s V. maximum potent ia l rate of carboxy lat ion, 0 J m a x l imi ted by J m a x umol m"^ s W , W, actual rates of carboxylation l imi ted by 2 J Rubisco, electron transport rate umol nf s W mean p r o f i l e s o i l water storage mm WUE canopy water use e f f i c i e n c y ra t io (=FC/E) g g - 1 , % z height above d m z M , z Q roughness lengths for momentum, heat mm z g depth of s o i l heat f lux plates (= 50) mm Z instrument height m - xxi -Greek Symbols a P r ies t ley -Tay lo r (1972) E/E g q a ' a applied to non- i r r igated vegetation B Bowen Ratio (H/LE) r adiabat ic lapse rate ( « 0.01) °C n f 1 r* C0 2 compensation point in absence of photorespirat ion Pa Y pyschrometric "constant" kPa K - 1 Y^ instrument value of Y kPa K - 1 Ac C0 2 concentration d i f f e r e n t i a l uL L - 1 m~1 Ae vapour pressure d i f f e r e n t i a l kPa AT, AT dry and wet bulb temperature d i f f e r e n t i a l s °C W Az sensing head separation distance m e M /M ( « 0.622) v a C z/1 Z natural decrease in e with a l t i t u d e Pa n f 1 8 so lar e levat ion angle degrees 9V s o i l volumetric water content m m K molecular d i f f u s i v i t y for heat in dry a i r mm2 s - 1 A energy of photosynthetic f i x a t i o n ( » 10.47) kJ g - 1 AF canopy net photosynthesis energy f lux ? density W m~d p density of dry a i r kg n f 3 t atmospheric transmission c o e f f i c i e n t Y n integral s t a b i l i t y influence function for sensible heat t ransfer Y M d iabat i c p r o f i l e correct ion function for momentum transfer Y g p r o f i l e to ta l s o i l water potent ial MPa - x x i i -McNaughton-Jarvis (1983) ^coupling factor" fl - [1 + v/((s + Y ) r A g c ) ] _ 1 fl appl ied to unir r igated vegetation surfaces - x x i i i -ACKNOWLEDGEMENTS I wish to extend my sincere thanks to my academic supervisor, Dr. Andy Black, for seven years of continuously wise and enthusiast ic support, for his pat ience, f r iendship and an a l l round good-natured working r e l a t i o n -sh ip . I a lso appreciate the patience of his long -suf fer ing wife and fami ly , who have always welcomed me at the i r home, even when Andy and I were d iscus -sing research related matters. I a lso wish to thank the other members of my supervisory commit-tee , Dr. T. M. B a l l a r d , Dr. D.L. Golding and Dr. J . P . Kimmins, for the i r advice and ass is tance , as well as Dr. J . Wilson, who always offered encourage-ment while working wonders in obtaining f inanc ia l support. Dr L.M. Lavkulich also provided encouragement and moral support, not to mention hundreds of computer d o l l a r s . I must a lso acknowledge the contr ibut ion of the U.B.C. Faculty of Graduate Studies who saw f i t to award me Univers i ty Graduate Fellowships for the period 1982-1985, and the Faculty of Forestry who awarded me Donald S. McPhee Graduate Fellowships in 1984 and 1985. Drs. J . Bassman and J . Crane both deserve recognit ion for gett ing 14 me interested and involved in the construction and use of the C0 2 equipment used in 1982, and for being e a r l i e r members of my supervisory committee. I a lso owe a lo t to Doug Beames, for his design, construction and maintenance of the f i e l d microcomputer and IRGA power suppl ies , not to mention many in te res t -ing discussions re la t ing to CMOS e lect ron ics and machine language programming. Barry Wong provided me with invaluable help in the arcane art of c a l l i n g FORTRAN subprograms from main programs wr i t ten in a more sensible language. Drs. Mike Novak and Dave Spitt lehouse have s i m i l a r l y maintained interest in my e f fo r t s over the years . Mike p a r t i c u l a r l y has always been w i l l i n g to take time out to discuss the f i n e r points of quantum mechanics, b lack -ho les , r e l a t i v i t y , superconductors and evapotranspiration theory (in order of preference) , while Dave has provided invaluable assistance in supporting our appl icat ions to the B.C. Min is t ry of Forests for research funding. - xxiv -I a lso wish to express my thanks to Mr. John Harwijne of Crown Forest Industr ies for cooperation and assistance at the Courtenay s i t e in 1981-1982 and to s t a f f of Woodlands Services D iv is ion of MacMillan Bloedel L t d . , p a r t i c u l a r l y but not exc lus i ve l y , Mr. Glen Dunsworth, for much needed cooperation, l o g i s t i c and f inanc ia l support at the Dunsmuir s i t e during 1983-1986. I must also express my grat i tude and appreciat ion to my long-suf fer ing g i r l - f r i e n d , Deb, for putt ing up with me and my e c c e n t r i c i t i e s for more than three years . Her charm and compassionate support have helped maintain my san i t y , even though I may not always have acknowledged i t at the time. She has also helped me on many occasions, in p a r t i c u l a r by entering most of my references into the computer, and by ca re fu l l y proofing the f i n a l draf t of t h i s t h e s i s . F i n a l l y , but by no means l e a s t , I must acknowledge the everpresent inf luence of my graduate student col leagues, past and present, for many hours of construct ive d iscuss ion , c r i t i c i s m and ass is tance. Frank K e l l i h e r , Marty Osberg, Don G i l e s , Nigel Livingston and Paul Sanborn a l l deserve special mention in that they helped with measurements and data c o l l e c t i o n at some time. Many others , too numerous to mention by name, in the Department of So i l Science, the Faculty of Forestry and down at the Graduate Student Centre, have affected my view of the world in some way. I t would also be t r a d i t i o n a l to acknowledge the e f fo r t s of a t y p i s t , except that there was none. Instead I would l i k e to recognise Intel and Z i log Corporations for developing microprocessors, and Microsoft Corporation for developing the software that enabled me to prepare the manu-s c r i p t myself , exact ly the way I wanted i t done. Therefore a l l typographic errors and omissions are e n t i r e l y my r e s p o n s i b i l i t y . . . . - 1 -INTRODUCTION The product iv i ty of a l l vegetative ecosystems i s l imi ted by water, energy, n u t r i t i o n and geomorphic fac to rs . So i l water and so lar energy are widely recognized to be of major importance in l i m i t i n g the growth of con i fe r -ous forest in the P a c i f i c Northwest, (e .g . Waring and Frankl in 1979). In th i s context, much research has been done at drought-prone s i t e s on eastern Vancouver Island by Black and coworkers, aimed at understanding the ef fects of seasonal s o i l water d e f i c i t s on canopy conductance, evaporation and product-i v i t y of young Douglas - f i r ( P s e u d o t s u g a menziesli (Mirb.) Franco) plantat ions (Black 1979; Black et a l . 1980; Tan e t a l . 1977, 1978). Work during 1981-1982 (Price e t a l . 1986; K e l l i h e r e t a l . 1986) suggested that removing sa la l understory vegetation from around the base of one of a pa i r of s i m i l a r - s i z e d trees resulted in small but s i g n i f i c a n t i n c r -eases in s o i l water p o t e n t i a l , stomatal conductance and photosynthesis rate for the treated t ree . However the small s t a t i s t i c a l sample on which these conclusions were based, made i t desirable to extend the study to a larger sca le . The study reported here was i n i t i a t e d in 1983 as part of a B r i t i s h Columbia Min is t ry of Forests Section 88 funded research project to invest igate the e f fec t of sa la l ( G a u l t A e r i a shallon Pursh.) vegetation on the water a v a i l a b i l i t y to and growth of plantat ions of young Doug las - f i r . The plan was to monitor the s i t e for one year and then t reat hal f by removing a l l the understory vegetation for a period of up to two years , using the Bowen Ratio/ energy balance approach to compare the di f ferences in canopy conductance, evaporation and photosynthesis rate for both treated and untreated p l o t s . - 2 -It soon became apparent the e f f o r t required to c lear large enough plots for r e l i a b l e energy balance measurements would be impossibly expensive, and that as o r i g i n a l l y envisaged, the project was overambitious. With the same overa l l object ives in mind, the research e f f o r t became concentrated in various areas: evaporation and photosynthesis from the whole stand (reported in t h i s study) ; the dependence of sa la l understory evaporation on avai lab le energy and atmospheric vapour pressure d e f i c i t (Osberg 1986); and an ongoing study of the e f fec t of sa la l on storage of s o i l water ava i lab le to the t rees , and consequent e f fects on product iv i ty (Black et al. 1984, 1985, 1986; Appendix X I ) . In attempting to resolve the complex interact ions of factors a f f -ect ing forest canopy photosynthesis i t became apparent that not a l l necessary information was ava i lab le from the resul ts of the f i e l d work. To be able to understand and predict long-term responses of the stand to seasonal changes in s o i l water storage, i t was decided that another major object ive of t h i s thesis would be to develop a mechanistic model ( i . e . a " l o w - l e v e l " model, fol lowing the terminology of Landsberg (1981, 1986)). I t was ant ic ipated that th is model should be used for test ing the hypothesis that sa la l removal would contribute to increased stand product i v i t y , and might u l t imately be used to help make management decisions re la t ing to vegetation c o n t r o l . More import-a n t l y , the model might enable the predict ion of annual forest product iv i ty from annual t ransp i rat ion est imates, along the l ines of simple y i e l d models proposed for a g r i c u l t u r a l crops by Tanner (1981) and Tanner and S i n c l a i r (1983). F a i l i n g these ob ject ives , i t should reveal where more or better information would be required to achieve them. Chapter I of t h i s thesis reports on the canopy energy balance and photosynthesis measurements, and presents observed diurnal and seasonal - 3 -re lat ionships between measured meteorological var iables and s o i l water storage and canopy conductance, evaporation and net photosynthesis ra tes . In Chapter II these re lat ionships are combined with physiological data obtained from the work of Farquhar and coworkers (e .g . Farquhar and Sharkey 1982; Caemmerer and Farquhar 1981; Kirschbaum and Farquhar 1984) to develop a model of canopy evaporation and photosynthesis. This model i s then tested using some of the weather and s o i l water storage data of Chapter I to estimate the measured evapotranspiration and photosynthesis rates . The qua l i t y of the model's predict ions i s d iscussed, and in p a r t i c u l a r , compared to that of Ja rv is et al. (1985). In the concluding chapter, the more general impl icat ions derived from Chapters I and II are reviewed in the l i g h t of the above d iscuss ion . REFERENCES Black, T.A. 1979. Evapotranspiration from Douglas f i r stands exposed to s o i l water d e f i c i t s . Water Resour. Res. 15: 164-170. B lack, T .A . , P.M. Osberg, D.T. Pr ice and D.G. G i l e s . 1986. Ef fects of sa la l removal on the water balance and tree water stress in young Douglas - f i r p lantat ions on the east coast of Vancouver Is land. MacMillan Bloedel , Woodlands Serv ices , unpublished report . Black, T .A . , D.T. P r i c e , F.M. K e l l i h e r and P.M. Osberg. 1984. Ef fect of understory removal on seasonal growth of a young Douglas - f i r stand. B.C.M.F. Forest Service Annual Report (1983-84). Black, T .A . , D.T. P r i c e , P.M. Osberg and D.G. G i l e s . 1985. Effect of reduction of sa la l competition on evapotranspiration and growth of early stage Doug las - f i r p lantat ions . B.C.M.F. Forest Service Annual Report (1984-85). B lack, T .A . , C .S . Tan and J . U . Nnyamah. 1980. Transpirat ion rate of Douglas-f i r trees in thinned and unthinned stands. Can. J . So i l S c i . 60: 625-631. Caemmerer, S . von, and G.D. Farquhar. 1981. Some re lat ionships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153: 376-387. - 4 -Farquhar, G.D. and T.D. Sharkey. 1982. Stomatal conductance and photosynthesis. Ann. Rev. Plant P h y s i o l . 33: 317-345. J a r v i s , P . G . , R.S. Miranda and R. I . Meutze l fe ldt . 1985. Modelling canopy exchanges of water vapor and carbon dioxide in coniferous forest p lantat ions . In B.A. Hutchinson and B.B. Hicks (eds . ) . The Forest-Atmosphere In teract ion . D. Reidel Publ ishing Company, pp 521-542. K e l l i h e r , F .M . , T.A. Black and D.T. P r i c e . 1986. Estimating the ef fects of understory removal from a Douglas f i r forest using a two-layer canopy evapotranspiration model. Water Resour. Res. 22: 1891-99. Kirschbaum, M.U.F. and G.D. Farquhar. 1984. Temperature dependence of whole-leaf photosynthesis in Eucalyptus pauc i f lo ra S ieb . ex Spreng. A u s t . J . Plant P h y s i o l . 11: 519-538. Landsberg, J . J . 1981. The number and qua l i t y of the dr i v ing var iables needed to model tree growth. Stud. For. Sue. 160: 43-50. Landsberg, J . J . 1986. The Physiological Ecology of Forest Production. Academic Press, London. 198 pages. Osberg, P.M. 1986. Lysimeter measurements of sa la l understory evapotrans-pi rat ion and forest s o i l evaporation a f te r sa la l removal in a Douglas-f i r p l a n t a t i o n . Unpubl. M.Sc. t h e s i s , Univers i ty of B r i t i s h Columbia, Department of So i l Science, Vancouver, B.C. P r i c e , D.T., T.A. Black and F.M. K e l l i h e r . 1986. Ef fects of sa la l understory removal on photosynthetic rate and stomatal conductance of young Doug las - f i r t rees . Can. J . For. Res. 16: 90-97. Tan, C . S . , T.A. Black and J . U . Nnyamah. 1977. Character i s t i cs of stomatal d i f f u s i o n resistance in a Douglas f i r forest exposed to s o i l water d e f i c i t s . Can. J . For. Res. 7: 595-604. Tan, C . S . , T.A. Black and J . U . Nnyamah. 1978. A simple d i f f u s i o n model of t ransp i ra t ion appl ied to a thinned Doug las - f i r stand. Ecology 59: 1221-1229. Tanner, C.B. 1981. Transpirat ion e f f i c i e n c y of potato. Agro. J . 73: 59-64. Tanner, C.B. and T.R. S i n c l a i r . 1983. E f f i c i e n t water use in crop product-i on : research or re-search? In H.M. Taylor , W.R. Jordan and T.R. S i n c l a i r (eds . ) . L imitat ions to E f f i c i e n t Water Use in Crop Production, pp 1-27. American Society of Agronomy, Crop Science Society of America, So i l Science Society of America, Madison, Wisconsin. Waring, R.H. and J . F . F r a n k l i n . 1979. Evergreen coniferous forests of the P a c i f i c Northwest. Science 204: 1380-1386. - 5 -CHAPTER_I MEASUREMENT_OF_ENERGY NII_£HOIOSYNIHESIS_OF_A_DOUG I__INIEODUCIION There i s much evidence to indicate that the growth rates of f o r -ests and a g r i c u l t u r a l crops are reduced by growing season s o i l water d e f i c i t s (e .g . Kozlowski 1982; Waring and Frankl in 1979; Ja rv is e t a l . 1976; Gholz et a l . 1976; Boyer 1976a, 1976b; Br ix 1962, 1979; Tanner 1981). In B r i t i s h Columbia, Black and coworkers (e .g . Tan and Black 1976; Tan e t a l . 1977; K e l l i h e r e t a l . 1986) found evidence of a causal re lat ionsh ip between reduced evaporation rates of Douglas - f i r [Psevdotsvga wenziesll Mirb . (Franco)) and the presence of sa la l (Gaultberla sballon Pursh) understory vegetation at a drought-prone s i t e on eastern Vancouver Is land. Tan e t a l . (1978) succeeded in estimating canopy level evaporation from a simple d i f f u s i o n model, using porometer-determined stomatal conductance charac te r i s t i cs related to s o i l water storage and weather condi t ions . Pr ice et a l . (1986; Appendix X) extended th is work to show that during a period of accumulating drought, decreasing s o i l water potent ia l resulted in reductions in both stomatal conductance and f o l i a r net photosynthesis ra te , and that the removal of sa la l understory ameliorated these e f f e c t s . They also made e s t i -mates of canopy photosynthesis (and water use e f f i c iency ) from these r e s u l t s , which were found consistent with dendrometer data obtained for pairs of s i m i -l a r trees with and without the understory removed. However these estimates - 6 -were treated with caut ion , because there i s abundant evidence demonstrating the serious d i f f i c u l t i e s in extrapolat ing measurements of leaf processes to the canopy level ( Jarv is and McNaughton 1986; Milne et a l . 1985; Leverenz et al. 1982; Schulze et al. 1977; Farquhar and Wong 1984; Baldocchi 1987). This approach has nevertheless met with some success. Apart from Tan et al. (1978), Whitehead et al. (1984) found only small d i f ferences in needle conductance of Pinus sylvestris both between current and one-year-old needles, and between d i f f e r e n t leve ls in the canopy. S i m i l a r l y , Troeng and Linder (1982b) found l i t t l e va r ia t ion in leaf net photosynthesis with pos i t ion in a canopy of 20-year -o ld P. sylvestrls, although in t h e i r case, needle conductance was much less consistent . Recognising these l i m i t a t i o n s , i t was considered essent ia l to study the re lat ionships between growing season s o i l water d e f i c i t s and forest growth at a larger scale than that of indiv idual leaves, i . e . at the canopy l e v e l , with the ult imate objective of t ry ing to determine the importance of moisture competition from understory vegetation in l i m i t i n g timber product-i v i t y . I t was also desirable to see whether any ef fec ts of decreasing s o i l water storage on photosynthesis could be observed on a large and s t a t i s t i c a l l y s i g n i f i c a n t number of t rees . In view of the l o g i s t i c d i f f i c u l t i e s experienced 14 in using the C0 2 technique and porometry to get meaningful resu l ts from even a s ingle pa i r of trees (see also Leverenz et al. 1982), i t was decided to adopt a micrometeorological method for measuring both canopy net photosyn-thesis rates and canopy conductance. The objectives of t h i s chapter are (1) to report the success of the experimental techniques, (2) to present and compare energy balance analyses for a se lect ion of days, with par t i cu la r regard to the re lat ionships between s o i l and weather condit ions and measured growth rates for both young and old stands of D o u g l a s - f i r , (3) to demonstrate - 7 -and discuss seasonal changes in canopy evaporation, photosynthesis and water-use e f f i c i e n c y (WUE) for wet (1983) and dry (1984) growing seasons. In the fo l lowing chapter, object ives (2) and (3) w i l l be considered in more de ta i l in the context of a canopy model based on the l e a f - l e v e l photosynthesis models of Farquhar and coworkers (e .g . Farquhar and Sharkey 1982; Kirschbaum and Farquhar 1984). I I SITE_DESCRIPTION_AND_EXPE^ The technique used to determine canopy net photosynthesis rate required accurate measurement of the small gradients in atmospheric carbon dioxide concentration above the canopy. This measurement i s most d i f f i c u l t to make over an aerodynamical l y very rough canopy such as that t y p i c a l l y found over t a l l , thinned coniferous f o r e s t . The thinned stand near Courtenay on eastern Vancouver Is land, where much previous work had been undertaken, was abandoned par t l y because of i t s height and aerodynamic roughness and part ly because during 1982 i t was discovered that a s i g n i f i c a n t proportion of Doug las - f i r trees at that s i t e were infected with fungal pathogens (notably laminated root ro t , Phslllnas welrll). With these factors in mind, a new s i t e , with a denser stand of shorter t rees , was selected in February-March 1983, which became the locat ion for research work undertaken during the 1983, 1984 and 1985 growing seasons. - 8 -S i t e _ D e s c r i r j t i o n The new s i t e (referred to as the "Dunsmuir Creek" s i te ) i s located at 49°02'N, 124°12'W, on eastern Vancouver Island in MacMi1lan-Bloedel's Chemainus Woodlands D i v i s i o n , approximately 30 km south west of Nanaimo, B.C. It has an elevat ion of 450 m with north -easter ly aspect varying between 2 and 5°. In the B.C. Biogeocl imatic C l a s s i f i c a t i o n system, the s i t e i s c l a s s i -f i e d as a t r a n s i t i o n between the CDFa2 (Nanaimo and Georgia Wetter Maritime Coastal Douglas - f i r ) and CWHa2 (East Vancouver Island Drier Maritime Coastal Western Hemlock) Variants (Kl inka 1977). The s o i l was c l a s s i f i e d as a Duric Dystr ic Brunisol intergrading to a Humo-Ferric Podzol (Sanborn 1984, personal communication). The parent material i s a compacted g l a c i a l t i l l over la in by a r e l a t i v e l y well drained surface layer of 0 . 8 - 1 . 0 m depth. For the surface layer of sandy-loam m a t e r i a l , the f ine bulk density was determined to be 900 kg m , while the volumetric stone f rac t ion was found to be approximately 50% (Black et al. 1985; Osberg 1986). S i l v i c u l t u r a l l y , the s i t e consists of two d i s t i n c t areas. The more westerly area of about 5 ha i s referred to as the "old stand", which was evidently planted to Douglas - f i r (Psevdotsuga menziesii (Mirb.) Franco) around 1963. Since then extensive natural regeneration of Doug las - f i r and other species has occurred, mainly of western hemlock {Tsuga heteropIiYlla (Raf.) S a r g . ) , with some western red cedar (Thuja plloata Donn) and small proportions of western white pine (Pluos montloola Dougl.) and lodgepole pine (Plnos oontorta var . latilolla Dougl . ) . Currently the stocking density varies con-s iderably but i s general ly higher than that of most commercial stands. This was one major reason for se lect ing the s i t e , since i t was ant ic ipated that the e f fec t i ve canopy roughness should be lower than for most stands in the region. - 9 -Nevertheless the stand height was found to be quite v a r i a b l e , general ly in the range 7-8 m, but with patches of poorer growth and occasional trees (notably white pine) up to 50% t a l l e r than average. Just to the east of the old stand was a second area of about 3 ha, where the stocking density was s i g n i f i c a n t l y lower and the trees younger, with an average height of about 1.0 m. This area was planted to Douglas - f i r around 1971, but the surv ival was poor and natural regeneration not s u b s t a n t i a l . For these reasons, the area was referred to as the "young stand". Both the old stand and the young stand had s i g n i f i c a n t ground-level shrub and herb vegetation (referred to as the "understory") , cons ist ing predominantly of s a l a l , but with associated species including Oregon grape {Berberis spp . ) , bracken (Pterldlvm aqulllxam) and grasses. Leaf area index, L A , was measured (see Appendix II) and found to be about 1.0 for the sa la l vegetation in the young stand, compared with 1.5 for the understory and 3.5 for the tree canopy in the old stand. L A for the trees in the young stand was estimated to be about 0 . 2 . ll±Z S t a n d _ N u t £ 2 t i o n a l _ S _ t a t u s On 1 December 1983, samples of three Doug las - f i r trees were selected for f o l i a r nutr ient a n a l y s i s . Six such samples were taken from the old stand and s i x from the young stand, g iv ing 36 trees in t o t a l . Analysis of the samples was undertaken by laboratory s t a f f at MacMi1lan-Bloedel Woodlands Services D iv is ion and the resul ts are given in Appendix IV. Of the 10 elements for which determinations were made, nitrogen was c l e a r l y d e f i c i e n t . Iron analys is i s known to be d i f f i c u l t , but i f the reported values are accurate, then i ron was probably also d e f i c i e n t . Concentrations of copper - 10 -and, in the old stand, of z i n c , may also have been l i m i t i n g (Bal lard 1987, personal communication). 11^3 Weather A basic automated weather s tat ion was set up in the young stand and maintained more or less continuously through the summers of 1983 (10 June-1 December), 1984 (14 Apri1-15 November) and 1985 (11 A p r i l - 9 November). Meteorological var iables monitored on an hourly basis were solar i r rad iance , a i r temperature and r e l a t i v e humidity (both measured at 1.3 m height ) , prec-i p i t a t i o n , wind ve loc i t y (1.3 m height ) , wind d i rec t ion (1.1 m height ) , and s o i l temperature at depths of 0, 200 and 500 mm. Sensor outputs were i n t e -grated using a datalogger (Campbell S c i e n t i f i c Inc . , Logan UT, model CR21), and recorded on cassette tape. Analysis was performed using a set of micro-computer programs developed for the purpose, and some pert inent observations are summarised here. The p reva i l ing synoptic scale westerly winds carry moist a i r from the P a c i f i c Ocean, which resul ts in frequent orographic r a i n f a l l over higher elevat ion central areas of Vancouver Is land. From observations made in 1983 and 1984 i t became evident that at the Dunsmuir Creek s i t e on c lear days, easter ly sea breezes from the Georgia S t r a i t (mean wind d i rec t ion about 80°) would normally develop during day-t ime. Conversely, on overcast and rainy days, day-time wind d i rec t ion general ly remained wester ly . The summer of 1983 proved to be r e l a t i v e l y wet: for the period 1 June to 1 September, tota l recorded p r e c i p i t a t i o n was 211 mm, compared with about 61 mm during the same period of 1984. In 1983, the longest period without appreciable p r e c i p i t a t i o n was only 42 days (16 July to 27 August), - 11 -during which there was approximately 15 mm t o t a l , compared to 66 days (29 June to 3 September) with only 21 mm in 1984. In 1984, the pyranometer f a i l e d resu l t ing in missing so lar i r r a d -iance data for the period 14 Apr i l to 21 June. These data were estimated using a regression equation determined from d a i l y to ta l s co l lected at Dunsmuir Creek in 1983, compared with da i l y to ta l s recorded at the Nanaimo weather s tat ion during the same per iod. With data for 168 days between 1 June and 1 December 1983, the regression equation obtained was: S D = 1.0729SN - 1.7389 ( r 2 = 0.83) (1) -2 -1 where S n , S N are d a i l y so lar i r radiance to ta l s (MJ m d ) at Dunsmuir Creek and the Nanaimo weather s tat ion respect ive ly . I I ..4 S o i l _ W a t e r _ S t o r a a e _ a n d _ P o t e n t As part of a related study, some of the sa la l understory vegetat-ion was to be removed as an experimental s i l v i c u l t u r a l treatment, so s o i l water storage measurements were i n i t i a t e d in June 1983, to attempt to charac-t e r i s e the s i t e water regime before any treatment had taken p lace. Five neutron s o i l moisture probe access tubes were i n s t a l l e d in the old stand and seven in the young stand. In A p r i l 1984, two 30 by 40 metre p lots were l a i d out in the old stand to study the ef fects of sa la l understory removal on the a v a i l a b i l i t y of water to the t rees . The treated p lot was located to the north and west of the meteorological tower (see I I .5 below) so that with the easter -l y winds normal for c lear days, the treatment would not a f fec t energy balance measurements over the whole stand. Conversely, the untreated control p lot was posit ioned d i r e c t l y east of the tower. Hence estimates of s o i l water content - 12 -in the old stand were based on a root-zone depth of 0.82 m, determined as the approximate mean depth of a l l locat ions where neutron probe access tubes were i n s t a l l e d in 1983 and in the untreated p lot in 1984. Because of the high stone content in the s u r f i c i a l layer , the neutron tubes were i n s t a l l e d by digging narrow trenches about 1.0 m in length. The removed s o i l was separated according to depth interva l and covered to minimise evaporation. Each trench was about 0.30 m wide but narrowed to a point at one end. A v e r t i c a l notch was made at the pointed end. Each tube was pushed f i rmly against the undisturbed s o i l of the notch, and the trench ca re f u l l y b a c k f i l l e d , tamping each layer to emulate the o r ig ina l bulk density as c lose ly as poss ib le . The surface organic layer was then replaced. Using t h i s i n s t a l l a t i o n method, i t was possible to ensure that a l l tubes extended completely to the underlying t i l l , but with the minimum amount of disturbed s o i l in the probe's "sphere of in f luence" . [For further discussion of th is technique see Osberg (1986).] F i e l d measurements made with the neutron probe (Campbell P a c i f i c Nuclear Hydroprobe Model 501) were used to determine average root-zone vo lu -metric s o i l water content (8 V). The probe was ca l ib rated in 1984 using g r a v i -metric data co l lected l o c a l l y . The equation obtained by regression of the neutron probe counts during 30 s against the measured gravimetric water content, a f t e r correct ing for the high volumetric stone f rac t ion was: ey = 0.217(N/N$) - 0.075 ( r 2 = 0.88) (2) where N and N g are the measurement counts and standard counts respect ively (Black et al. 1985, 1986; Osberg 1986). N g was determined with the probe in i t s s h i e l d , located upright over an access tube. Measurements were made general ly at two week i n t e r v a l s , though for Ju ly and August 1983, measurements - 13 -were made weekly. For each access tube, measurements were made in 0.15 m depth increments s ta r t ing at 0.15 m, and extending to the bottom of the tube. Water retention charac te r i s t i cs were found to be typ ica l for a coarse textured s o i l . Saturation volumetric water content (equal to the s o i l porosity) was about 29%. At " f i e l d capacity" ( » - 0 . 0 1 MPa), the volumetric water content (8 V) at both 0.1 and 0.4 m depth was around 15%, while the " w i l t i n g point" value ( « - 1 . 5 MPa) was about 4%. Consequently the s o i l ' s volumetric "ava i lab le water content" was estimated to be about 11%, which when mu l t ip l i ed by the mean root-zone depth of 0.82 m indicated a maximum avai lab le root-zone water storage of 90 mm. The neutron hydroprobe measurements of 8 y were mu l t ip l i ed by the root-zone depth to ca lcu late p r o f i l e average tota l s o i l water potent ia l (W). These per iodic measurements of W were combined with recorded weather data to estimate s o i l water storage on intervening days. Dai ly evaporative losses between each measurement of W were assumed to be proportional to the da i l y to ta l S (since R n was not measured for the whole season), while drainage and surface runoff were assumed to be zero. Total r a i n f a l l recorded during any day was added to the estimated i n i t i a l W for the fo l lowing day. In conjunction with the measurements of W, sets of s ix ca l ibrated s o i l thermocouple psychrometers (Wescor Inc . , Logan UT, model PCT-55) were i n s t a l l e d to measure s o i l water potent ia l ( Y s ) . In June 1983 one set was i n s t a l l e d at representative locat ions in both the old stand and the young stand. Two more sets were i n s t a l l e d in the old stand in Ju ly 1984, one in each of the two experimental p l o t s . At each l o c a t i o n , pai rs of psychrometers were placed approximately 0.45 m apart , at depths of 0 .15 , 0.45 and 0.75 m. Readings were made at least once per week through July and August 1984. - 14 -Af ter consideration of both the laboratory-determined water reten-t ion data and the f i e l d measurements of s o i l water storage and p o t e n t i a l , a curve of best f i t was determined using an equation of the form suggested by Campbell (1985): f . - -0.00323 (W/W.)"3-3 2 (3) where Ys i s in MPa, W i s the root zone water storage and Wg i s W at satur -at ion ( « 238 mm at 8 v = 29%), both assuming 0.82 m rooting depth. Both values of W and ¥ estimated from (3) are reported in the d iscuss ion . I I ..5 Enerf ly__Balance/Bowen I I . 5 . 1 Measurement System The micrometeorological approach adopted was the well establ ished energy balance/Bowen Ratio method (Black and McNaughton 1971; Spitt lehouse and Black 1979, 1980; Aston 1985b). However, several important modif icat ions were made to enable measurement of the canopy C0 2 concentration gradient and to promote easier data reduction. Analogous techniques have been used by Jarv is et al. (1976), Leuning et al. (1982). Verma et al. (1986) used an eddy-cor re la t ion approach to achieve s i m i l a r measurements over a deciduous fo res t . Ver t i ca l temperature and vapour pressure gradients were measured using a reversing d i f f e r e n t i a l psychrometric apparatus (Spitt lehouse and Black 1980). For the old stand i t was mounted on an 8.5 m meteorological tower. FD300 s i l i c o n diodes were used as dry and wet bulb temperature sensors, l o c -ated in r e f l e c t i v e radiat ion shie lds (termed the "heads") at the ends of two tubular "arms". The v e r t i c a l separation of the sensors was set at 3.0 m, with the lower head at a height of 7.8 m, i . e . jus t above the tops of the t a l l e s t - 15 -trees nearby and below the top of the roughness sublayer. For the young stand, where less turbulent mixing was to be expected, the head separation was 1.0 m, while the distance between the vegetation and the bottom sensor was approximately 0.2 m. Fetch was about 300 m in the easter ly (prevai l ing wind) d i r e c t i o n for the old stand and about 100 m east and west for the young stand. In the case of the old stand, the fetch distance was less than ideal since with an assumed zero plane displacement of 0.7 canopy heights (Appendix I I I ) , the fetch ra t io was about 50. Af ter consideration of the work of Jarv is et al. (1976) and Gash (1986) and of a ser ies of ca lcu lat ions of expected adjustment under various atmospheric s t a b i l i t i e s (Ayotte 1984, personal communication), i t was concluded the upwind distance was s u f f i c i e n t to allow adjustment of a i r a r r i v i n g at the psychrometric sensors. Furthermore, independent estimates of da i l y canopy evaporation, derived from measurements of W in both young and old stands, indicated that as a i r crossed the boundary only modest adjustment in evaporation rate would occur. Therefore any advect-ive e f fec t from the young stand would be un l ike l y to ser ious ly influence energy balance measurements made over the old stand. (See also III below, Appendix XI and Black et al. 1985, 1986.) The l imi ted fetch made i t necessary to place the sensors in the roughness sublayer, rather than the so -ca l led "constant f lux layer" further above i t . It i s recognised that th i s i s not des i rab le , because the roughness elements can cause large eddies where the v e r t i c a l concentration gradients are steep and v a r i a b l e , resu l t ing in poor measurements of the mean f luxes . How-ever, the work of Raupach and Legg (1984) indicates that the Bowen Ratio tech-nique should be acceptable when used with in the roughness sublayer of closed canopies (with small distances between the roughness elements), and provided the source/sink d i s t r i b u t i o n s for the e n t i t i e s being measured are s i m i l a r . - 16 -Denmead and Bradley (1985) made Bowen Ratio measurements of evapotranspiration in the roughness sublayer of a coniferous canopy which compared favourably with eddy cor re la t ion measurements made higher up. Vent i la t ion of the sensors was achieved using a 24 VDC radial fan (Pamotor, model RL90-18/24). The heads were reversed every 15 minutes and 5 minutes allowed for sensor e q u i l i b r a t i o n immediately a f t e r reversa l . Sensor voltages were then integrated for the las t 10 minutes of each 15 minute per iod , with a sampling frequency of approximately 2 Hz. Average dry and wet bulb gradients (AT and ATW) were determined every 30 minutes, by ca lcu lat ing the mean of the integrated values of AT and AT obtained in the two 10 minute w integrat ion per iods. The net canopy C0 2 f lux density (F c ) was determined by measuring the d i f ference in atmospheric C0 2 concentration (Ac) between the two heads using an i n f r a - r e d gas analyser ("IRGA", described in d e t a i l below) operating in d i f f e r e n t i a l mode. A i r over the canopy was sampled continuously through two intakes located next to the psychrometer heads, and with the same ver t i ca l separat ion, A z . The intakes were made of copper tubing to minimise C0 2 adsorption on the inner surfaces, and designed to prevent rainwater from enter ing. Each intake was connected to a short length of strong f l e x i b l e Tygon tubing, so that reversing of the heads would not be obstructed. On the mast supporting the reversing system, each of the two lengths of Tygon tubing was joined to 35 m lengths of 8.5 mm I.D. tubing ("Dekabon" from Eaton Corp. , OH). The Dekabon has very low absorpt i v i t y for C0 2 ( Jarv is 1983, personal communication), but in order to compensate for i t s small c r o s s - s e c t i o n , each tube ac tua l l y consisted of two p a r a l l e l lengths joined by "T"-connectors at both ends to al low a volumetric flow rate double that of a s ingle tube. - 17 -A i r was drawn along the tubes by two series-connected 24 VDC fans (Pamotor, model RL90-18/24), which generated approximately 300 Pa of suct ion. A t h i r d "T" - junct ion connected the paired tubes, so that the fans exerted suction equal ly on both intakes and dif ferences between t h e i r flow rates were minimised. The flow rates were monitored using a water manometer connected between the "T" - junct ion (upstream of the fans) and the atmosphere. The mano-meter was in turn ca l ib rated using a "wet - test" type gas flowmeter. Since the resistance of the flowmeter was not i n s i g n i f i c a n t , i t was necessary to c a l i b -rate the manometer with the flowmeter using a small 120 VAC vacuum pump (Gast) to obtain a flow rate representative of that obtained in the f i e l d . Within the normal operating voltage range, the flow rate of a i r through the reversing system was found to vary l i n e a r l y with (and therefore could be calculated from) the voltage applied to the fans. With an appl ied 24.5 VDC, the flow rate through each intake was found to be approximately 6.5 L m i n - 1 at S .T .P . Before entering the "T" - junct ion at the fans , the a i r from each intake was well mixed in a 4 L glass j a r using a low powered 12 VDC rotary fan (Micronel ) . Af ter ex i t ing from the mixing vesse ls , each of the two a i r streams was continuously subsampled v ia a small copper "T" - junct ion , using one of the IRGA's two internal sampling pumps. Before entering the IRGA, each subsample a i r stream was dr ied by passing i t through a column of D r ie r i te contained in a copper cy l inder of approximately 100 mL volume. It was found that the D r i e r i t e would las t approximately 36 hours before needing replace-ment. In p r a c t i c e , i t was replaced rout inely once per day. The IRGA was an ADC laboratory model (Analyt ic Development Co. , Hoddesdon, U.K., Series 225 Mk 2 "Plant Physiology") modified for f i e l d use. The AC transformer was disconnected, and a DC power supply constructed to provide the necessary voltages (+15 V, -15 V and +27 V). This DC power supply - 18 -operated from seven 12 V car b a t t e r i e s . In a d d i t i o n , the IRGA's internal sample pumps and the CG^ purge pump required 120 VAC 60 Hz power which was provided by a low-power, high e f f i c i e n c y , 12 VDC inverter b u i l t in the labor -atory . The IRGA was encased in an insulated wooden box so that the power required to maintain the thermostatted heat t rac ing system would be minimised. In p r a c t i c e , i t was found necessary to remove the i n s u l a t i o n , otherwise the IRGA tended to overheat on warm days. Af ter drying and f i l t e r i n g , the two subsample a i r streams entered the IRGA's analys is and reference c e l l s , and the d i f f e r e n t i a l volumetric C0 2 concentration measured. Flow rates were maintained at approximately 200 mL m i n - 1 , while the pressure d i f ference between the two c e l l s was monitored rout inely with a water manometer and maintained at less than 10 mm. The a i r intakes reversed with the psychrometric heads every 15 minutes, and a s i m i l a r 5 minute e q u i l i b r a t i o n period was allowed before voltage output from the IRGA was integrated. However, to allow for the time required for a i r to flow from the intakes to the IRGA (and hence synchronise the measurements of sensible and latent heat f luxes with those of F c ) , i t was found necessary to delay the integrat ion cycle for Ac by two minutes, r e l a t i v e to AT and AT , w This i s explained in more d e t a i l below. In addit ion to AT and AT W , the ha l f -hour l y mean absolute dry and wet bulb temperatures (T, T ) were measured using one of the psychrometric heads. A solarimeter (Kipp and Zonen, model CM-5) mounted on the tower at 7 . 5 m height , was used to measure the so lar i r radiance (S) , while a net radio -meter (Swissteco, type S - l ) was used to measure net i r radiance ( R n ) « So i l heat f lux density at 50 mm depth (G 5) was recorded using a pa i r of s o i l heat f l u x plates (Middleton). Heat storage in the top 50 mm layer of s o i l was c a l -culated from the ha l f -hour l y change in mean s o i l temperature ( T s ) , measured - 19 -with a pa i r of v e r t i c a l l y - i n t e g r a t i n g diode thermometers (Tang et a l . 1974). Atmospheric pressure, P a , was measured d a i l y using an aneroid barometer a ca l ib rated against a mercury unit in the laboratory. I I . 5 . 2 System C o n t r o l and Data C o l l e c t i o n Reversing of the psychrometers and a i r intakes, timing of the integrat ion cyc les , integrat ion of the sensor output voltages, ca lcu la t ion of secondary quant i t ies and data recording were a l l achieved using a f i e l d micro-computer designed and constructed in the laboratory. The f i e l d microcomputer used low power CMOS c i r c u i t r y based on an RCA CDP 1802 microprocessor with EPROM-based software. Time and date were recorded continuously using an on-board clock/calendar with battery backup. Every ha l f -hour , (on the hour and hal f -hour) the complete timing sequence was i n i t i a t e d with head reversa l . At t h i s po in t , integrat ion of S , R n and G g for the previous hal f -hour ceased, and commenced for the current ha l f -hour . Af ter a two minute delay to allow a i r to a r r i ve from the intakes, integrat ion of Ac ceased and ca lcu lat ions were done on a l l data co l lec ted in the previous ha l f -hour . These c a l c u l a t i o n s , together with the basic sensor output data , were pr inted out by a small p r in ter housed in the computer enclosure. Five minutes a f te r reversa l , integrat ion of the diode voltages (T, Tw, AT and AT ) commenced. Seven minutes a f te r reve rsa l , integrat ion of the W IRGA output voltage (Ac) commenced. At 15 minutes, integrat ion of T, T , AT and ATW ceased and the heads reversed, followed two minutes la te r by a halt to integrat ion of Ac. At 20 minutes integrat ion of T, T , AT and AT recom-W W menced, followed two minutes la te r by Ac . At 30 minutes, the heads reversed again, and integra ls of a l l sensor voltages, other than the IRGA output, were stored temporari ly . Two minutes into the beginning of the next ha l f -hour , in tegrat ion of A c ceased, and the hal -hour ly energy balance calcu lat ions - 20 -completed. The ent i re cycle already described was then repeated. In addit ion to pr inted output, a l l data were recorded on cassette tape for subsequent mainframe computer a n a l y s i s . I t was found in 1983 that night - t ime measurements in the young stand tended to overrange due to the strong temperature and C0 2 inversions that would form in s t i l l a i r under c lear s k i e s . This problem was resolved in 1984 by inser t ing a voltage overrange detection routine in the microcomputer software, so that the e f fec t i ve s e n s i t i v i t y on these channels could be halved whenever overranging occurred. I I .5.3 System Checks and C a l i b r a t i o n Every day the system was in operat ion, a complete set of routine checks was undertaken. In p a r t i c u l a r , the psychrometer readings were checked against an Assmann psychrometer held at the midpoint between the heads, and the wet bulb reservoirs were checked and topped up as necessary. The IRGA was ca l ibrated usual ly once per day using a standard gas (Matheson) of known volumetric C0 2 concentration (339 uL L - 1 ) using both the IRGA manufacturer's c a l i b r a t i o n procedure and the method of Parkinson and Legg (1978), the l a t t e r enabling an estimate of ambient atmospheric C0 2 concen-t ra t ion to be made. Using these c a l i b r a t i o n methods, errors in the assumed absolute concentration of the reference gas have r e l a t i v e l y l i t t l e e f fect on accuracy when the IRGA i s used in d i f f e r e n t i a l mode. In 1983 the d i f f e r e n t i a l s e n s i t i v i t y of the analyser was set at 0.81 mV (uL L - 1 ) - 1 , but a f te r imple-menting the voltage overrange detection rout ines , the IRGA s e n s i t i v i t y was increased in 1984 to 2.0 mV (uL L " 1 ) - 1 . In both cases the s e n s i t i v i t y was found to vary by no more than 2.5%. The s e n s i t i v i t y was par t l y dependent on barometric pressure, so measurements of atmospheric pressure were made rout-- 21 -ine ly using an aneroid barometer, which was la te r ca l ib rated against a high qua l i t y mercury barometer. IRGA zero d r i f t was of the order of 1 uL L - 1 per day which was not considered a serious problem since the head reversal every 15 minutes e f f e c t i v e l y removed any s i g n i f i c a n t error t h i s might cause. With integrat ion of the Ac voltage signal for 20 minutes every ha l f -hour , i t was found that the overal l resolut ion for the measurement of Ac (across e i ther head separation distance) was general ly better than 0.02 uL L " 1 . Since day-time Ac over the forest with a 3.0 m separation was general ly found to be of order 0.5 uL L - 1 , the resolut ion (of ±47.) was considered adequate. In operat ion, the complete measurement system required fourteen 12 VDC vehic le b a t t e r i e s . I t was found possible to jus t maintain the system with overnight charging of two b a t t e r i e s . This corresponded to an overal l power consumption of approximately 75 watts. I_l±6 C a l c u l a t i o n s I I . 6 . 1 E q u a t i o n s Used The system determined each of the terms in the energy balance equation for the stand: Rn = H + LE + G Q + M + AF C (4) where R n i s the net i r rad iance , H and LE are the sensible and latent heat f lux d e n s i t i e s , G Q i s the surface s o i l heat f lux densi ty , M i s the change in canopy heat storage expressed on a ground area basis and A F C i s the energy f lux density associated with photosynthetic f i x a t i o n . - 22 -Signal voltages for the nine input channels were each mul t ip l ied by an appropriate c a l i b r a t i o n factor to give the basic data from which the remaining var iables were ca lcu la ted . The Bowen Rat io , B, was i n i t i a l l y c a l c -ulated from: H P " LE " 1 + (e - 2 ) e P a (AT - rAz) Y (Ae - ZAz) ( 5 ) where y i s the psychrometric constant, corrected for barometric pressure and temperature (see Appendix I ) . The v e r t i c a l vapour pressure d i f ference across the psychrometric heads, A e , i s given by Ae = ( s w + Y < ) A T - Y ^ T where y^ (= 1.03y) i s the instrument value of Y . S i s the slope of the saturation W vapour pressure curve (de /dT) at T and AT, AT are measured over Az. Hal f -hourly mean vapour pressure, e, was calculated from the psychrometer equation, e = e w - Y < ( T - T w ) , where e w i s the saturat ion vapour pressure at T . The adiabat ic temperature gradient (lapse rate) r , has the value - 0 . 0 1 " C m " 1 , while Z i s the corresponding natural decrease in e assuming perfect mixing, obtained from Z = (e/PJAP /Az = -eg/(R T„ ) « -0.0001167e rn"1 where P, i s the a a a Is a measured barometric pressure (Thorn 1975; Mcintosh and Thorn 1969). The term (1 + (e - 2)e/P f l) where e i s the ra t io of the molecular masses of water vapour to dry a i r (M„/Ma « 0 .622) , i s a small correct ion factor which allows for the V a small v e r t i c a l wind component a t t r ibutab le to the sensible heat f lux density , H, a f fec t ing the measured value of the latent heat f lux densi ty , LE, as discussed in Webb et a l . (1980) and Leuning et a l . (1982). For the purposes of these c a l c u l a t i o n s , s i m i l a r i t y of the eddy d i f f u s i v i t i e s for heat, water vapour and C0 2 (K^, K v , K c ) was assumed for a l l atmospheric s t a b i l i t i e s . - 23 -The surface s o i l heat f lux density (G Q) was determined from: G Q = G 5 + dH s/dt (6) where: dH./dt = z (AT /At)C (7) d and H $ i s the heat storage per unit ground area, in the s o i l layer of depth z $ , (above the heat f lux p l a t e ) , AT g /At i s the ha l f -hour l y change in average temperature for t h i s layer , and C $ i s the s o i l volumetric heat capacity of th i s layer determined from per iodic gravimetric s o i l samples. The rate of heat storage by the old stand canopy, M, was estimated from: M " M Ha + M L a + MHb <8) where M H a and M L a are the rates of sensible and latent heat storage in the canopy a i r , and M „ b i s the rate of sensible heat storage in the canopy biomass. These terms were determined using a i r temperature and humidity measurements made wi th in and above the canopy, the average height of the canopy, and the average stand basal area estimated from mensurational data, using the approach of Stewart and Thorn (1973), a lso reported in Thorn (1975), Ja rv i s et a l . (1976) and McCaughey (1985). The energy being f ixed in canopy photosynthesis, A F , was calculated as fo l lows . The ra t io AFC/H i s given by: AF. - M r M P a - e)Ac C = — ^ = £ ( 9 ) H M a c p L " P a " e ( 1 " e ) 3 ( ^ - TAz) -where M c i s the molecular mass of CO^ (44.01 g m o l - 1 ) , A i s the energy of photosynthetic f i x a t i o n (* 10.47 kJ g - 1 [ C 0 2 ] , Denmead (1969)) and c p i s the - 24 -s p e c i f i c heat of dry a i r (taken as 1.003 kJ k g - 1 K - 1 ) . The small correct ion term (P, - e)/[P - e ( l - e ) ] accounts for the a i r samples being dried before a a entering the analyser , so that Ac i s expressed with respect to dry a i r at constant temperature, assuming the a i r samples were brought to the same temperature and pressure inside the IRGA measurement c e l l s (Webb et a l . 1980, Leuning et a l . 1982). By combining (4) , (5) and (9) the fo l lowing equations can be derived (Monteith 1973): LE = (R n - G Q - M)/(l + B + A) (10) and: A F c = A ( R n " G o " M ) / ( 1 + P + A ) ( H ) where A = pC = AF C /LE. The rate of canopy t ransp i rat ion (mg[H20] nf 2 s - 1 ) was calculated from E = LE/L where L, the latent heat of vaporisation of water, was corrected for temperature (see Appendix I ) . S i m i l a r l y , the net - 2 - 1 canopy mass f lux density of C 0 2 , (ug[C0 23 m s ) , was calculated from F £ = AF C /A. The water use e f f i c i e n c y mass ra t io (WUE) was then calculated as F c /E. The eddy d i f f u s i v i t y for heat t ransfer was also estimated using: = ^ (12) M C (AT - TAz) a where C f l i s the volumetric heat capacity of dry a i r and H i s taken from (4). Hal f -hour ly average wind ve loc i t y data recorded by the automated weather s tat ion were used to estimate average wind ve loc i t y every half -hour for both the young and old stands and then merged with the energy balance data to enable ca lcu la t ion of canopy aerodynamic resistance (Appendix I I I ) . Wind ve loc i t y data above the old stand were estimated from weather s tat ion measure-- 25 -merits made during the study per iod, using a re lat ionship obtained from anemo-meter data recorded by the weather s tat ion and a sens i t i ve Casel la anemometer i n s t a l l e d above the old stand canopy for 10 days during September 1983. Canopy aerodynamic resistance to heat and mass d i f f u s i o n , r A , was calculated from the hourly mean wind ve loc i t y approximately corrected for the Monin-Obukhov s t a b i l i t y function (Dyer and Hicks 1970; Webb 1970; Thorn et al. 1975) using an i t e r a t i v e approach described in Appendix I I I . Having deter -mined r^, i t was possible to obtain the bulk canopy resistance to water vapour d i f f u s i o n , r , from the Penman-Monteith equation (Monteith 1965): pc D r c = — ^ + r [ ( B S / Y ) - 1] (13) C Y L E M * where D i s recorded at the measurement height and s i s de /dT evaluated at T. The procedure used to determine the canopy aerodynamic resistance term also y ie lded a value for the eddy d i f f u s i v i t y for momentum t rans fe r , <M , at the measurement height which could be compared with the K H calculated from (12). At night - t ime and during ear ly morning, when net ava i lab le energy was small ( i . e . when (R n - G - M) was close to zero) , combined with low wind v e l -oc i t y and a strong inversion over the canopy, (12) could occasional ly y i e l d negative values for K M . Since to a good approximation K„ = KM in stable condit ions (Thorn 1975), F c could be recalculated when was negative using: KMAc(P - e) F c = ~ ~ (14) AzR T„ C K where R c i s the gas constant for C0 2 (0.1889 J g " 1 K f 1 ) . - 26 -Canopy conductance for water vapour, g , was calculated as the reciprocal of r c , but to estimate the internal C0 2 concentrat ion, c^, g c was converted to the canopy conductance for C 0 2 , g c ' , by d i v id ing by the rat io D v /D c , where D y and D c are the molecular d i f fus ion c o e f f i c i e n t s for water vapour and C0 2 in a i r respect ive ly , and D v/D c » 1.65 at 20 °C (Monteith 1973). Hence the canopy i n t e r c e l l u l a r C0 2 concentration at the e f fec t i ve mean sink l e v e l , c.., was obtained from: c i • WW • w <15> _3 where c f l i s the ambient C0 2 concentration in g m assumed constant throughout the canopy, and corrected for ambient temperature and pressure by the term T,P a/(TVP.) where T. i s the i ce -po in t temperature (273.16 K) and P 0 i s s tan-I a F> 5 1 S dard atmospheric pressure (101.325 kPa). Caution should be exercised in ident i f y ing c. with the actual mesophyll C0 2 concentrat ion, mainly because the value of g 1 was not derived from measurements on ind iv idual leaves. In p a r t i c u l a r , the possible contr ibut ion of s o i l evaporation to canopy LE was not considered, although the work of K e l l i h e r et al. (1986) on a more open canopy suggests t h i s component would not be large . The assumptions of constant D, T and c f l used in (13) and (15) which imply perfect mixing with in the canopy, are also supported by measurements of T and D made below and above the canopy during 1984. Nevertheless, the analys is of Finnigan and Raupach (1987) shows that there could well be large errors in g 1 resu l t ing from t h i s approach. As an index of the evaporation rate r e l a t i v e to the equi l ibr ium rate , E e q , day-time and 24 hour values of the ra t io E/E e q (termed a ' ) were calculated from: a ' = E/E = LE(s + Y)/[s(R n - G Q - M)] (16) - 27 -where the terms LE(s + Y ) and s(R n - G Q - M) were the t o t a l s of ha l f -hour ly values ca lcu lated over the appropriate per iod . P r i e s t l e y and Taylor (1972) o r i g i n a l l y i d e n t i f i e d E/E as the parameter a appl icable to non - l imi t ing e g s o i l moisture condi t ions . S i m i l a r l y , to invest igate the re la t i ve contr ibut ions of ava i lab le energy and D in inf luencing LE, values of fl (McNaughton and Jarv is 1983; Ja rv i s 1985a, Ja rv is 1985b), were calculated rout inely using: fl = (1 + s/ Y )/ ( l + S/Y + l / r A g c ) (17) Even though fl has often been applied to "wel1-watered" vegetat ion, i t was considered a useful index of the degree of "decoupling" between ambient D and LE for stressed condit ions as w e l l , represented here by fl'. Hal f -hour ly values of fl1 were observed to be unre l iab le when ava i lab le energy was smal l , so fl ' was recorded as the mean of f i ve ha l f -hour l y measurements recorded between 10:45 and 12:45 PST. This fol lows Verma et al. (1986) who c i t e midday values for fl. Appendix V gives a l i s t i n g of the Pascal/VS program wri t ten to do these c a l c u l a t i o n s . The algorithms are e s s e n t i a l l y the same as those contain-ed in the microcomputer ROM code, except for (7) and (13) through (17). The ca lcu lat ions of canopy heat storage were' added to improve accuracy of the energy balance determination, but were not essent ia l for f i e l d monitoring. I I .6.2 Data R e d u c t i o n Energy balance data were co l lected on a to ta l of 73 days, 24 dur-ing 1983 and 59 during 1984. Prel iminary analys is consisted of p lo t t ing 24 hour graphs of key var iables including A T d , AT W , Ac , 8 , , K M , g £ and D and 24 hour energy balance diagrams. Examination of these p lots allowed - 28 -e l iminat ion of those days when substantial amounts of day-time data were miss ing , and when there were obvious technical problems with the energy balance system. Subsequent analys is resulted in the e l iminat ion of a further 16 days' data co l lected over the old stand. The resul ts presented here are based on deta i led analys is of 6 days' data from the young stand col lected between 4 and 13 August 1983, and of 32 days from the old stand, the l a t t e r consist ing of 6 days co l lected between 18 August and 1 September 1983, and 26 days between 18 July and 30 August 1984. Subsequent data manipulation included the ed i t ing of c lear l y spurious points (notably those which occurred during maintenance work on the system or while c a l i b r a t i n g the IRGA) and interpo lat ion of missing data when these could be predicted with confidence. Detai led analys is of da i l y trends reported below was however res t r i c ted to eight days where the data were subject to l i t t l e or no a l t e r a t i o n . For days where more s i g n i f i c a n t changes were required, or where delet ion of some records was unavoidable, data were used only to estimate 24 hour means and t o t a l s . The majority of days studied were characterised by intermittent cloud cover, rather than being completely c lear or completely overcast. Since c lear days were the easiest to in te rp re t , the se lect ion of days for detai led analys is was l i m i t e d . Nevertheless, on days with s i m i l a r weather and s o i l condi t ions , canopy responses were also general ly s i m i l a r . For example, energy balance data for 20 August 1983 presented here showed s i m i l a r trends to those for 25 August 1983 reported in Black et al (1984). - 29 -I H RISULTS_AND_DISCySSION The canopy Bowen Ratio/energy balance/photosynthesis measurement system, which as mentioned e a r l i e r ( I I .5 .3) was found to have a resolut ion of about +0.02 uL L - 1 , was capable of measuring canopy C0 2 f l ux dens i t ies around - 2 - 1 10 ug m s , in t yp ica l day-time turbulent condi t ions . However, measurement of night - t ime C0 2 gradients often proved d i f f i c u l t , because the IRGA would overrange when large Ac occurred in s t i l l condi t ions , so that determinations of canopy dark resp i rat ion rates were not always considered trustworthy. Consequently in terpretat ion of these resul ts was res t r i c ted mainly to day-time measurements. The day-to-day changes in W (estimated using the procedure des-cribed in I I .4) were compared with d a i l y to ta l s of E computed from the energy balance/Bowen Ratio measurements. For the young stand, d a i l y measured E was general ly about 20% higher than that indicated by estimated d a i l y changes in s o i l water storage (AW). However, for the old stand the discrepancy was much greater , with d a i l y measured E often 50 to 100% higher than AW. Af ter careful consideration i t i s thought that there are probably s i g n i f i c a n t errors in the estimate of W. The volumetric stone f rac t ion used to es tab l i sh the neutron probe c a l i b r a t i o n was based on only one measurement and could be 20% too high. The root-zone depth, estimated as the average depth of a l l access tubes i n s t a l l e d on the s i t e , could be 20% too shallow when appl ied to the old stand. It i s possible that some water was ava i lab le to plant roots from the under-l y ing t i l l , or from drainage flow at the base of the s u r f i c i a l layers . The untreated p l o t , immediately east of the tower, was found to be d r i e r than most other areas further upwind, yet hal f the neutron probe access tubes were concentrated in t h i s p l o t . ( It was thought t h i s area would contribute more to - 30 -the energy balance measurements at the tower, so that the measurements of W should be weighted accordingly , but in retrospect th i s was probably a mistake.) Taken together, the above possible sources of error could explain the apparent discrepancy in the separate ca lcu lat ions of evaporation. Hence there are l i k e l y to be errors in the absolute values of W presented in the fo l lowing d iscuss ion , but they should not a l t e r the conclusions concerning seasonal changes in W and the i r e f fects on canopy processes. Lindroth (1984) had even worse problems obtaining agreement between s o i l water balance and energy balance estimates of forest evaporation. He attempted to explain i t in terms of n o n - s i m i l a r i t y between K and K u over the f o r e s t , but i t i s much more V n l i k e l y that estimates of W were in error since only two access tubes were used. The eight days for which deta i led resu l ts are to be discussed and compared are presented in three sect ions . The f i r s t sect ion discusses two pairs of c lear days over the old stand, which c l e a r l y show the interact ing ef fects of s o i l water potent ia l and atmospheric vapour pressure d e f i c i t on canopy conductance. In the context of the McNaughton-Jarvis Q, the second section compares typ ica l diurnal energy balances for both stands during c lear and cloudy days, before examining the apparent seasonal changes in canopy evaporation as the s o i l water potent ia l gradually decreases. The th i rd section discusses the ef fects of weather and s o i l water storage on diurnal and seasonal canopy net photosynthesis rates measured above the old stand. - 31 -1 1 I ^ i E f f e c t s _ o f _ S o l l _ W a t e r _ P o t e n t 2 6 l i c l t _ o n _ E v a p . o t ran sp_i r a t i o n_R 0 1 d _ S t a n d During the 1983 growing season, average W in the old stand ( e s t i -mated from measurements at s i x neutron probe access tubes i n s t a l l e d that year) was always s i g n i f i c a n t l y higher than at any time during the 1984 growing season a f te r measurements begin in mid - Ju ly . For th i s reason a c lear day in 1983, (20 August, F i g . l . l a ) was compared with three days in 1984 (F igs . 1.1b-l . l d ) which exhibited s i m i l a r c lear sky condit ions but progressively lower W. I I I . 1 . 1 S o i l Water P o t e n t i a l Apart from s o i l water storage, condit ions on 20 August 1983 and 20 August 1984 were remarkably s i m i l a r in weather terms (F ig . 1.2a) , so that e f fec ts of high and low Y $ on canopy conductance (and net photosynthesis) could be compared, even e l iminat ing leaf age as a possible cause of v a r i a t i o n . On 20 August 1983, when Y $ was high ( » -0 .035 MPa; W » 128 mm) evaporation climbed to a maximum around noon, and decreased s tead i l y afterward. Calculated g £ remained around 7-8 mm s " 1 a l l morning and only decreased s l i g h t l y during the afternoon as D reached a maximum of about 1.6 kPa around 16:00 PST. Therefore, on t h i s day D was evidently the major influence over E, because under these condit ions i t had l i t t l e e f fec t on g c > A year l a t e r , on 20 August 1984, Y s had decreased to about - 1 . 1 5 MPa (W « 51 mm) while maximum D reached only 1.5 kPa. However, E, and hence g £ , was general ly about 20% less a l l day than had been observed in 1983. Since the diurnal va r ia t ion in D was very s i m i l a r on both days, the reduction in g £ on 20 August 1984 must be a t t r ibu tab le to the s i g n i f i c a n t l y lower W. - 32 --200-1 1 1 1 1 1 1 1 1 1 1 1 1 0 4 8 12 16 20 24 HOURS (PST) Figure l . l a . Canopy energy balance for the old stand on 20 August 1983. Plotted variables are the solar irradiance (S), net Irradiance (R n), the sum of canopy heat storage and soil surface heat flux densities (GQ + M) and the flux densities of sensible heat (H), latent heat (LE) and photosynthetic fixation energy (AFC). First of four selected clear days showing effect of decreasing W on canopy LE. Soil water content (W) was about 128 mm, profile average soil water potential (¥ ) was about -0.035 MPa and maximum S was 840 W nf 2 at 12:15. - 33 --200 H 1 1 1 1 1 -\ 1 1 1 1 1  0 4 8 12 16 20 24 HOURS (PST) Figure 1.1b. Canopy energy balance for the old stand on 24 Ju ly 1984. W was about 84 mm, Y g was about -0 .17 MPa and maximum S was 876 W n f 2 at 12:15. [Notation as in F i g . 1 .1a] . - 34 -Figure 1.1c. Canopy energy balance for the old stand on 29 Ju ly 1984. W was about 81 mm, Y was about - 0 . 2 MPa and maximum S was 882 W n f 2 at 12:15. [Notation as in F i g . 1 .1a] , - 35 -LU -200 -H 1 1 1 1 1 1 1 1 1 1 1  0 4 8 12 16 20 24 HOURS (PST) Figure l . l d . Canopy energy balance for the old stand on 20 August 1984. W was about 51 mm, ¥ s was about - 1 . 1 5 MPa and maximum S was 829 W nf 2 at 12:15. [Notation as in F i g . 1 .1a] . - 36 -4 - i 1 i i 1 1 1 1 1 1 1 i HOURS (PST) Figure 1.2a. Comparison of canopy evaporation rate (E) in response to atmo-spheric saturation vapour pressure def ic i t (D) for two identical clear days over the old stand, with very different W. During the 1983 growing season, W remained high and on 20 August was about 128 mm, whereas in 1984, W declined steadily to about 51 mm by 20 August. - 37 -Figure 1.2b. Comparison of canopy evaporation rate (E) in res-ponse to atmospheric saturat ion vapour pressure d e f i c i t (D) for two c lear days over the old stand during the 1984 growing season with essent ia l l y s i m i l a r W. On 24 July W was about 84 mm, decreasing to 81 mm on 29 J u l y , the small d i f ference being due to 5.6 mm p r e c i p i t a t i o n on 25 Ju l y . - 38 -I I I . 1 . 2 Vapour P r e s s u r e D e f i c i t Clear sky solar i r radiances (S) on 24 and 29 July 1984 were very s i m i l a r (F igs . 1.1b and 1.1c) , but very high T, reaching 31 °C on 24 Ju l y , resulted in maximum D of almost 3.5 kPa at mid-afternoon compared to 1.75 kPa on 29 Ju ly (F ig . 1.2b). So i l water storage leve ls were intermediate between those of 20 August 1983 and 20 August 1984, being estimated at 84 and 81 mm on 24 and 29 July respect ive ly . High S contributed to r e l a t i v e l y high da i l y evaporation during t h i s period (generally around 3.5 mm per day), but th i s was somewhat compensated by 5.6 mm prec ip i ta t ion on 25 July (see Appendix VI I ) , resu l t ing in a comparatively small change in W. Hence average p r o f i l e Y g changed only s l i g h t l y between 24 and 29 July 1984, probably remaining at about - 0 . 2 MPa. In view of the evident moderate stomatal s e n s i t i v i t y to Y s observed above, t h i s small change i s not l i k e l y to have exerted a major inf luence on g c . Moreover, g £ on 24 July was only s l i g h t l y lower than on 20 August 1983 which suggests Y g was not low enough to have much ef fect in any case. Hence condit ions on 24 and 29 July enabled d i rec t comparison of the ef fects of high and low D on g £ and E. F i g . 1.2b shows that compared to 29 J u l y , E was s i m i l a r or only s l i g h t l y higher on 24 J u l y , even though D (considered representative of the vapour pressure d i f ference across the stomata) was general ly twice as large a l l day. Calculated values of g c were therefore general ly about half of those for 29 Ju ly and hence i t i s c lear that under these condit ions of s o i l water storage, g decreased s i g n i f i c a n t l y in response to high D. Therefore E was regulated mainly by the e f fect of D on the stomata. The responsiveness of the stomata to changes in D, and the re la t i ve i n s e n s i t i v i t y to changes in Y (at least in the range of potent ia ls at which measurements were made) are - 39 -subs tant ia l l y in agreement with the stomatal resistance charac te r i s t i cs deter -mined by Tan e t a l . (1977) for Douglas - f i r at the ind iv idual needle l e v e l . Values of g c observed on 29 July were s l i g h t l y higher than those on 20 August 1983, even though Y was lower and D higher. This apparent con-t r a d i c t i o n might be explained by a reduction in maximum g c occurring over the three week interva l between 29 July and 20 August 1984. Although th is i s somewhat speculat ive , there i s much evidence that maximum g w i l l decrease as the growing season progresses. Contr ibuting factors could include hardening of new shoots, c u t i c l e development on new f o l i a g e , needle f a l l of older f o l -iage, onset of dormancy in response to shorter daylengths (and possibly in response to low s o i l water potent ia l ) and other phys io logical reasons related to leaf ageing (e .g . Lindroth 1986; McMurtrie e t a l . 1986; Beadle e t a l . 1982). 111^2 Comrja r i s o n _ o f _ D l u rna Vap_ou r _ P r e s s u r e _ D e f 2 c l t _ a n d _ E v Stand_Canop_ie s McNaughton and Jarv is (1983) f i r s t introduced the concept of fl to indicate the degree of "decoupling" between E and D in the ambient a i r f low over an unstressed vegetation canopy (see (24) above; Ja rv is 1985a, 1985b). Evaporation from low- ly ing vegetation cover should be c h a r a c t e r i s t i c a l l y l inked to the ava i lab le energy (R n - G Q - M), because the aerodynamic r e s i s t -ance term ( r A ) over such vegetation i s r e l a t i v e l y h igh, while canopy ("surface") resistance ( 9 C _ 1 ) tends to remain low, causing the l e a f - a i r vapour pressure d i f ference to be depressed. By comparison, evaporation from t a l l e r forest vegetation w i l l be driven mainly by the imposed atmospheric vapour pressure d e f i c i t , because low aerodynamic and boundary layer r e s i s t -- 40 -ances ( p a r t i c u l a r l y in the case of coniferous forests) necessar i ly resul t in the leaves being subjected to atmospheric D approaching that of the mixed layer ( leaf temperatures w i l l be s i m i l a r l y wel l -coupled to that of the atmosphere). Strong stomatal s e n s i t i v i t y to D ( typical of coniferous forest vegetation) w i l l a lso contribute to canopy evaporation rates r e l a t i v e l y inde-pendent of i r rad iance . Conversely, since the young stand consisted predomin-ant ly of low broadleaved herb and shrub species, with general ly higher r A , and since the stomata of many broadleaved species are also less sens i t ive to changes in D, i t would be expected to exh ib i t a stronger dependence of LE on R n (higher fi) than seen over the old stand. I I I . 2 . 1 C o u p l i n g i n the O ld S t a n d on C l e a r Days F i g . 1.1 shows a c lear seasonal trend in the dependence of LE on R n . On 20 August 1983 (F ig . l . l a ) , when W was high, LE reached a maximum around noon (even ignoring the s l i g h t peak at 13:00 PST), and was therefore r e l a t i v e l y well correlated to the sinusoidal change in R n , with midday values of fi' in the range 0 . 1 9 - 0 . 2 2 . I t should be noted however that some cloud developed in the late afternoon, and while H responded strongly to f l u c t u -at ions in S and R n , the course of LE was large ly unaffected. The fo l lowing year, during Ju l y , even though W and Y S were subst-a n t i a l l y lower, values of fi' were very comparable to those on 20 August 1983, being in the range 0.16-0.23 on 24 July and 0.19-0.25 on 29 July (F igs . 1.1b and 1.1c) . On 24 J u l y , lower D resulted in less stomatal closure and a higher cor re la t ion between R n and LE than on 29 July (F ig . 1.2b). However by 20 August 1984, Y was much lower and, as noted above, evidently contributed to lower g c > Consequently, fi' was also s i g n i f i c a n t l y lower, in the range 0 .15-0 .17 , ind icat ing stronger coupling between D and LE. F i g . 1.Id i l l u s t r a t e s - 41 -t h i s very c l e a r l y since LE exhib i ts almost l i near decl ine through the day with l i t t l e re la t ionsh ip to R n . I I I . 2 . 2 C o u p l i n g i n the Young S t a n d On 9 August in the young stand (F ig . 1.3a) , c lear sky conditions p reva i led , whereas on 11 August (F ig . 1.3b), the sky was in termi t tent l y cloudy throughout the day. Day-time wind d i rec t ion was easter ly for 9 August but westerly for 11 August, which though hardly i d e a l , i s not considered a serious objection to comparing the data. So i l water storage and potent ia l were very high at t h i s t ime, decreasing only s l i g h t l y between the two days (from 198 to 196 mm; Y s » 0.0 MPa). A close re lat ionship between LE and ava i lab le energy i s c l e a r l y evident on both days (F igs . 1.3a and 1.3b) with f luc tuat ion in LE on 11 August being d i r e c t l y a t t r ibu tab le to cloudiness causing greater va r ia t ion in S and R n . Midday values of P.1 were in the range 0 .3 -0 .35 on 9 August and 0 . 3 - 0 . 5 on 11 August, i . e . s i g n i f i c a n t l y higher than those values normally observed over the old stand. However, the reduction in E between 9 and 11 August i s proport ionately greater than that in R n , which indicates that dependence of E on R n i s not complete and that D does a f fec t E (F ig . 1 .4) . This e f fec t i s seen in the higher g £ observed on 11 August, evidently responding to lower D. Evaporation rates from the young stand were general ly lower than from the old stand, shown by day-time B in the range 1 .4 -1 .8 on 9 August, and even higher on 11 August, compared to 1 .0 -1 .4 in the older stand (with lower W) a few days l a t e r . The main explanation for t h i s must be the low L A of around 1.0 measured in the young stand, in addit ion to some stomatal closure in response to high D. It was also noted that the growth habit of the sa la l was very d i f f e r e n t from that occurring in the shade of the old stand. The leaves were general ly smal ler , more v e r t i c a l l y oriented and f e l t more b r i t t l e , - 42 -HOURS (PST) Figure 1.3a. Canopy energy balance of the young stand on 9 August 1983, a c lear day. W was about 198 mm, ^ was close to zero and maximum S was 855 W m~l at 12:15. [Notation as in F i g . 1 .1a] . - 43 -HOURS (PST) Figure 1.3b. Canopy energy balance of the young stand on 11 August 1983, a cloudy day. W as about 196 mm, was about 0.0 MPa and maximum S was 824 W m" at 14:45. [Notation as in F i g . 1 .1a] . - 44 -CO E C O E E o D) 2-0 100 50 20 10 4 0 i 1 1 1 1 1 1 1 1 1 p • 9 August 1984 • 11 August 1984 2* 8 12 16 20 HOURS (PST) 24 Figure 1.4. Comparison of canopy evaporation rate (E) in response to atmo-spheric saturat ion vapour pressure d e f i c i t (D) over the young stand in 1983 on 9 August (c lear) and 11 August (cloudy) . Total canopy conduct-ance, g c , was estimated from E, D and estimated canopy aerodynamic res is tance , r^, using the Penman-Monteith combination equation. - 45 -(perhaps ind icat ing a th icker c u t i c l e ) . These features would contribute to lower leaf conductance to water vapour and greater heat t ransfer through the leaf boundary layer . Low canopy L A was also related to the f a i r l y high G Q (maxima exceeding 50 W m ) , recorded on both days which were correlated with R n , p a r t i c u l a r l y on 11 August (see F i g . 1 .3 ; note M was assumed to be zero for the young stand). By comparison, the old stand canopy had s i g n i f i c a n t l y smaller G Q , and although canopy heat storage, M, became a more s i g n i f i c a n t term, the _ 2 sum (GQ + M) was general ly smal ler , reaching midday maxima of about 40 W m _2 in Ju ly and only 30 W m in August. The d a i l y ra t io of M/GQ in the old stand was quite v a r i a b l e , but averaged about 0.2 for the days studied. I I I . 2 . 3 E f f e c t s o f C l o u d Cover on C o u p l i n g i n the O ld S t a n d When compared to energy balances for the old stand (F ig . 1 .1) , F igs . 1.3a and 1.3b suggest that a canopy with higher o.1 w i l l exh ib i t stronger cor re la t ion between f luctuat ions in R n and LE, p a r t i c u l a r l y those caused by cloudy condi t ions . By comparison, evaporation over the old stand should be less dependent on f luctuat ions in R R , with lower 0. ' . To test th is hypothesis, energy balances over the old stand were examined for two cloudy days, 26 August 1983 and 6 August 1984 when W was about 121 mm (¥ - -0 .045 MPa) and 72 mm (¥ « - 0 . 3 2 MPa) respect ive ly . Both days received approxi -mately 18 MJ m tota l so lar i r radiance (F ig . 1.5) and the sums of (GQ + M) were a lso smaller compared to c lear days over the old stand. Wind d i rect ions were predominantly easter ly on both days, though more var iable than on most c lear days. In sp i te of these s i m i l a r i t i e s , 2.8 mm r a i n f a l l on the previous day and drying winds of about 2.0 m s - 1 during the night probably contributed to lower T and higher vapour pressure on 6 August 1984, so that maximum D was jus t under 1.0 kPa compared with about 1.5 kPa on 26 August 1983 (F ig . 1 .6) . - 46 -L U -200 H 1 1 1 1 1 1 1 1 1 1 1  0 4 8 12 16 20 24 HOURS (PST) Figure 1.5a. Canopy energy balance for the old stand on 26 August 1983. F i r s t of a pa i r of cloudy days showing e f fec t of decreasing W on canopy LE. W was about 121 mm, Y s was about -0 .045 MPa and maximum S was 825 W m at 11:45. [Notation as in F i g . l . l a ] . - 47 -800 i i i i i i i i i i i i -200 H 1 1 1 1 1 1 1 1 1 1 1 ! 0 4 8 12 16 20 24 HOURS (PST) Figure 1.5b. Canopy energy balance for the old stand on 6 August 1984. Second of a pa i r of cloudy days showing e f fec t of decreasing W on canopy LE. W was about 72 mm, T was about - 0 . 3 2 MPa and maximum S was 702 W m~2 at 12:15. [Notation as in F i g . 1 .1a] . - 48 -4 i i i i 1 1 1 1 1 1 1 i HOURS (PST) Figure 1.6. Comparison of canopy evaporation rate (E) in response to atmo-spheric saturation vapour pressure def ic i t (D) for two cloudy days over the old stand, with very different W. On 26 August 1983 W was about 121 mm, whereas on 6 August 1984 W was about 72 mm. - 49 -This d i f ference probably accounts for the s i g n i f i c a n t l y greater g c observed on 6 August 1984 (since lower W evidently had l i t t l e e f f e c t ) , which in turn allowed comparable, or even marginal ly greater , evaporation rates . Both cloudy days exhibited s l i g h t response of LE to changes in R n (more noticeable on 6 August 1984), but large decreases in R n during 26 August 1983 (F ig . 1.5a) produced temporary decl ines in D, which were coupled with drops in E (F ig . 1 .6 ) . Hence coupling between D and LE was never as strong as on the c lear days of 1984 (F ig . 1.2b), and fl1 was higher, in the range 0.20-0.30 on 26 August 1983 and 0.25-0.34 on 6 August 1984, approaching values observed over the young stand. I t i s possible that the higher P.' on 6 August 1984 was due in part to exaggerated g c resu l t ing from incomplete drying of the canopy (pa r t i c -u l a r l y the understory) , but the low morning values seen in F i g . 1.6 indicate that such an e f fec t could not have been very s i g n i f i c a n t . The work of Osberg (1986), undertaken at the Dunsmuir Creek s i t e in 1984 and 1985, showed that r a i n f a l l intercepted on the surface of the sa la l leaves would amount to about 0.1 mm. Assuming water accumulation on the Douglas - f i r needles to be s i m i l a r , the wetted canopy would carry a maximum of 0.5 mm. From inspection of the -2 -1 data , evaporation averaged about 25 mg m s between 14:00 PST on 5 August and 02:00 PST on 6 August ( F ig . 1 .6 ) , y i e l d i n g a to ta l of about 1.0 mm after the r a i n f a l l stopped, which strongly suggests the canopy should have been dry by dawn on 6 August 1984. However, Beadle et al. (1985b) reported unpub-l i shed data of Leverenz that show spuriously high g s may occur due to incomplete drying of stems and branches even when the fo l iage i s dry. In general cloudiness was correlated with lower a i r temperatures and hence lower D, resu l t ing in general ly higher g (at least through the range of ¥ reported here.) For t h i s reason the ra t io r J r . was lower and the 5 I* ft - 50 -ca lcu lated fi' higher than on c lear days with s i m i l a r W. In spi te of these observations, fi' remained lower than for the young stand. A very dramatic demonstration of the weak cor re la t ion between ava i lab le energy and LE in the old stand occurred on 4 August 1984 (Appendix IX) . The morning was per fect l y c lear but around noon, heavy cloud developed rapid ly and pers isted for the rest of the day. Within an hour both S and R n had decreased by a factor of 3 , while p dropped from about 2.0 to 0 .4 . The rapid decrease in ava i lab le energy therefore resulted in substant ial changes to H, but LE was v i r t u a l l y unchanged. I I I . 2 . 4 E f f e c t s o f S e a s o n a l Changes i n S o i l Water S t o r a g e on a ' and ft' I t should be possible to consider la te August 1983 as being s i m i -l a r (at least hydro logical ly ) to the ear ly part of a typ ica l "wet" growing season when W was higher. Assuming t h i s to be t rue , seasonal courses of W and midday P.' were p lotted for the three measurement per iods, i . e . ear ly August 1983 ( in the young stand), late August 1983 and July-August 1984 ( in the old stand) as shown in F i g . 1 .7 , together with day-time values of a 1 ( i . e . the Pr ies t ley -Tay lo r a parameter applied to s o i l water l imi ted vegetat ion) . C lear l y a ' and fi1 are c lose ly cor re la ted , except that a ' over the old stand was general ly higher than over the young stand, while P 1 was gener-a l l y lower. Normally the canopy evaporation from both young and old stands was at less than the equi l ibr ium rate (a 1 < 1) , the exceptions being when the canopies were wet. For the young stand day-time a ' was in the range 0 .8 -0 .9 which may be explained by the low L A and evident anatomical adaptations of the sa la l referred to e a r l i e r ( I I I .1 .2) contr ibut ing to high surface r e s i s t -ance. For the old stand day-time a 1 was general ly 0.8 or lower and decreased to 0 . 5 - 0 . 6 toward the end of the 1984 growing season, evidently in response to - 51 -Figure 1.7. Seasonal changes in day-time a ' ( s o l i d l ine ) and midday fl1 (dashed l ine ) related to W for the young stand in ear ly August 1983 (a) , the old stand in la te August 1983 (b) and the old stand during Ju l y -August 1984 (c ) . For the purposes of t h i s study, a ' and P.' are defined as the P r ies t ley - Tay lo r a and McNaughton-Jarvis fl parameters applied to vegetated surfaces for varying root-zone s o i l water storage. - 52 -decreasing Y $ . By comparison, midday P' was more conservative and decreased less markedly. Jarvis and McNaughton (1986, Table 1) give values for fi over a complete range of "unstressed" vegetation types. Agricultural crops generally l i e in the range 0 .6 -0 .9 , which indicates that evaporation is dependent mainly on irradiance while there is relatively l i t t l e stomatal response to ambient D. (Omega of 1.0 would indicate evaporation at the equilibrium rate, as would a of 1.0.) For comparison, Table 1.1 summarises values of fi (or Q 1) repres-entative of "natural" vegetative surfaces ( i . e . subject only to extensive management practices and not a r t i f i c i a l l y i r r igated) . The young stand, with i t s predominant cover of s a l a l , resembles the heather [Calluna vulgaris) moorland ecosystem studied by Miranda et al. (1984) in Scotland. However, the s imi lar i ty in fi hides some important differences, because i t is essent-i a l l y dependent only on the ratio r c / r A . Both resistances are lower for the heather moor, because l_A was larger (about 4.0 compared with 1.0 in the young stand at Dunsmuir Creek) and secondly r A was smaller probably because the Scottish s i te was more exposed and generally windier. Verma et al. (1986) also obtained similar values for P , when working over deciduous forest in Tennessee. In their case, r was lower presumably because, compared to conifers, the leaves of broadleaved tree species tend to have higher stomatal conductances, and are less responsive to increases in D, while r A was lower because of the greater canopy height and relative aerodynamic roughness. Coniferous forest has the both the lowest r A (or highest 9 A _ 1 ) and the greatest stomatal sensit iv i ty to D. However, the values of 0.15-0.25 for the old stand canopy are s l ight ly higher than for other coniferous forests, which may be attributed to higher r A than those reported elsewhere (e.g. Jarvis et al. 1976; McNaughton and Black 1973). Higher r. may be due Table 1.1 Approximate aerodynamic and canopy conductances, and derived midday values of McNaughton-JarvIs fl at Dunsmuir Creek compared with t yp ica l values for a select ion of vegetation canopies in the l i t e r a t u r e . Conductance units are mm s . Values of ft were calculated assuming T of 20 °C. Vegetation canopy 9 C 9 A n Source of data Dunsmuir "young stand" 5 25 0.38 This study 1 Heather moor 8 40 0.38 Miranda e t a l . (1984) 2 Dunsmuir "old stand" 6-11 100 0.15-0.25 3 This study Pine forest 7 170 0.11-0.15 Whitehead e t a l . (1984) Deciduous forest 10 50 0.38 Verma e t a l . (1986) Agr icu l tura l crops 20 ? 50 ? 0 . 4 - 0 . 9 Various Notes: Low leaf area Index of 1.0 leads to low g . c 2 Leaf area Index = 4.0. Probably an exposed windy location. The authors report P. to be around 0.32, but this 1s not a midday value. 3 Short dense forest (7-8 ra canopy height, with about 5000 stems per ha). 4 See Jarvis and McNaughton (1986) Table 2 for details. - 54 -to the rather lower tree height and higher stand density at the Dunsmuir Creek s i t e , which was selected par t l y for these reasons. So i l water content i s c l e a r l y important in determining fi, because comparison between "wet" and "dry" condit ions (20 August 1983 and 20 August 1984 respect ively ) shows that LE i s more c lose ly related to ava i lab le energy in the former case (see F igs . 1.1a and l . l d ) . Hence i t may be possible for the energy balance of a forest with high s o i l water storage to tend toward that of an a g r i c u l t u r a l crop. Table 2 of Ja rv i s and McNaughton (1986) gives values of fl for "unstressed vegetation canopies", but since most forests are not i r r i g a t e d , W for an "unstressed" forest w i l l t y p i c a l l y be lower (often below f i e l d capacity) than for an ag r i cu l tu ra l crop. Since i t i s well establ ished that g can be s i g n i f i c a n t l y influenced by Y (amongst other f a c t -o r s ) , and i f we assume r A in (17) to be e s s e n t i a l l y constant (even though, as Jarv is and McNaughton point out, i t may not be measured c o r r e c t l y ! ) , i t i s inev i tab le that P. for forests (and p a r t i c u l a r l y coniferous forests) can change appreciably with t ime. This leads to the conclusion that forests with except-i ona l l y high s o i l water storage w i l l exh ib i t A' values higher than those norm-a l l y expected. Higher than normal values of fl1 were also observed on the two cloudy days, which shows that low D may st imulate stomatal opening, and possibly a regional scale feed-forward response. 11X^3 P h y . s l c a l _ F a c t o r s _ I n f x^  Resp_ 2 i a t 2 o n_in_ t h e _ 0 1 d _ S t a n F i g s . 1.1 and 1.5 show that the f lux density of energy associated with net photosynthesis at the top of the canopy, A F C , was a very small comp-_ 2 onent of the overal l canopy energy balance, t y p i c a l l y less than 5 Wm dur-ing day-t ime. However, to maintain consistency between micrometeorology and plant physiology, comparison of the ef fects of s o i l and meteorological - 55 -variables on canopy photosynthesis and respiration rates are discussed in terms of C02 mass flux density. I I I .3.1 E f f e c t s of A i r Temperature Normally a broad diurnal correlation between g and F was obser-ved, both reaching maxima in early morning and declining through the rest of the day. However, i t was not i n i t i a l l y clear whether F£ was limited directly by effects of temperature (and other factors) or through the effect of g c decreasing in response to increasing D (directly related to increasing T). Leverenz (1981), Beadle e t a l . (1985b) and Price e t al. (1986) a l l concluded that day-time F was normally l ight - l imi ted , and that g would only become l imit ing to F c at high D. As w i l l be seen, these conclusions are substant-i a l l y confirmed in this study. F ig. 1.8 compares the diurnal variation in F c and c i for 24 and 29 July 1984, plotted together with a i r temperature, T, and the quantum flux density at the top of the canopy, Q, derived from Q = 2.04S, where 2.04 is an empirical factor determined by Meek e t al. (1984), and S and Q are in units - 2 - 2 - 1 of kW m and mmol m s respectively. In i t ia l examination might suggest that 24 hour net photosynthesis is negative. However, there is good reason to suppose that many of the night-time measurements of F c are spurious, due to a combination of low night-time d i f fus iv i t ies and low available energy which can both make accurate determinations of small fluxes extremely d i f f i c u l t . Fur-thermore, i t is necessary to allow for the contribution of the soi l respira-tory efflux of C02 to the canopy (R s ) . Measurements of R$ were made using the soda lime technique of Monteith e t al. (1964). A clear dependence of R$ on T or Tg was not detected (the results of Monteith e t a l . for an agricult -ural f i e l d were only s l ight ly clearer) , but Rg was found typical ly in the - 2 - 1 range 100-200 ug m s , suff icient to make 24-hour F generally positive. - 56 -o j" \ — 1 — i — 1 — i — 1 — r " i 0 4 8 12 16 20 24 HOURS (PST) Figure 1 .8 . Comparison of net canopy photosynthesis rates (F £ ) and estimated canopy i n t e r c e l l u l a r C0 2 concentration (c^) 1n response to quantum f lux density (Q) and a i r temperature (T) fo r two c lear days over the old stand during the 1984 growing season, with e s s e n t i a l l y s i m i l a r W On 24 July W was about 84 mm, decreasing to 81 mm on 29 J u l y , the small change being due to 5.6 mm p r e c i p i t a t i o n on 25 J u l y . The horizontal dotted l i n e p lot ted at -100 ug m s " 1 on the graph of F £ indicates the prob-able contr ibut ion of s o i l respiratory CO, to F . Table 1.2 Water use e f f i c iency (WUE) rat ios at the Dunsmuir Creek s i t e 1n 1983 and 1984 calculated from F_/E and expressed on a mass percentage b a s i s . DATE Daytime WUE Daytime WUE 24 hour WUE 24 hour WUE (uncorrected) (corrected for (uncorrected) (corrected for respirat ion) resp i rat ion) 9 Aug 1983 2 +0.2 - - 0 . 1 3 -11 Aug 1983 2 +0.5 - +2.01 -20 Aug 1983 +0.17 +0.38 - 0 . 6 9 - 0 . 3 6 26 Aug 1983 +0.30 +0.52 +0.361 +0.821 24 Jul 1984 - 0 . 1 2 +0.026 - 0 . 8 3 - 0 . 6 1 29 Jul 1984 +0.023 +0.17 - 0 . 4 9 - 0 . 2 6 6 Aug 1984 +0.49 +0.69 +0.741 +1.061 20 Aug 1984 -0 .013 +0.20 - 0 . 8 5 - 0 . 4 9 Notes: Values are obviously 1n error because the 24 hour WUE exceeds the daytime value, Indicating a downward CO^  flux at night. 2 Measurements on 9 and 11 August 1983 apply to the young stand, where soil respiration rates were not measured. Other dates apply to the old stand. - 58 -However, Edwards and So i l ins (1973) concluded that the soda-lime technique tends to s i g n i f i c a n t l y underestimate s o i l resp i rat ion rates , p a r t i c u l a r l y at temperatures above 20 °C. On the basis of his r e s u l t s , the true rates could - 2 - 1 be in the range 150-300 ug m s . Following Monteith (1973), l ines have been plotted on the graphs of F (F igs . 1 .8 , 1.9 and 1.10) to show the estimated contr ibut ion from s o i l resp i ra t ion . ) A i r temperatures were except ional ly high on 24 July 1984, and consequently F c was only barely pos i t i ve for a short period in the ear ly morning and s i g n i f i c a n t l y negative for the remainder of the day. High E also occurred on t h i s day (F igs . 1.1b and 1.2b) so 24 hour WUE was consequently very negative (Table 1.2) . By comparison, 29 July 1984 was much cooler and F £ higher than on 24 July so that 24 hour WUE was much less negative (F ig . 1.8 and Table 1 .2) . These observations are strong evidence that T has a major inf luence over F c . The reason i s l i k e l y to be increased resp i rat ion more than reduced carboxylation (Jones 1983). Kirschbaum and Farquhar (1984) modelled the temperature depend-ence of photosynthesis in E u c a l y p t u s pauoiflora and showed that i t would decrease at temperatures above 30 °C because of the increasing contr ibut ion of non-photorespiratory r e s p i r a t i o n , R d , and a progressive upward s h i f t in the C0 2 compensation po in t , r * . Da Costa e t al. (1986) reviewing work on forest s o i l s , considered the ef fects of a i r temperature and Y s on resp i rat ion in an a g r i c u l t u r a l crop, and found that with Y $ < - 1 . 2 MPa resp i rat ion decreased. In view of the general ly higher water potent ia ls observed here, i t i s probable that the only s i g n i f i c a n t factor inf luencing resp i ra t ion was a i r temperature. (Chapter 2 w i l l discuss possible functional re lat ionships between T and F c . ) There can be l i t t l e doubt that on most days (at least during July and August) a substant ial proportion of day-time net photosynthesis was lost through r e s p i r a t i o n . Increased photorespiration rates at high temperatures in - 59 -forest canopies were also observed by Beadle et al. (1985a) working in a stand of Plnus sylvestrls, while Yoda (1967, 1971 and 1978 c i ted in Ninomiya and Hozumi 1981), observed high night-t ime resp i rat ion rates in response to high a i r temperatures. The Dunsmuir Creek s i t e i s recognised to be of low product iv i ty and loss of photosynthate due to high resp i rat ion rates on hot days would be a contr ibut ing fac to r . Ja rv i s and Leverenz (1983, Table 8.7) present comparative annual carbon balances for three coniferous species. Even though many of the values are approximate, i t i s evident that for r e l a t i v e l y productive s i t e s , a subst-an t ia l proportion (70% or more) of annual photosynthesis (after deducting f o l i a r respi rat ion) i s los t through n o n - f o l i a r t issue resp i rat ion and morta l -i t y . For the lowest product iv i ty s i t e (0.6 tonnes h a - 1 y _ 1 ) these losses accounted for about 80% of annual photosynthesis. This could have serious impl icat ions for future temperate-zone coni fer forest product iv i ty throughout the wor ld , since projected increases in mean temperature (due to global increases in atmospheric C0 2 concentration) may resu l t in respiratory losses comparable to the ant ic ipated increases in photosynthesis. I I I . 3 . 2 E f f e c t o f Y on Canopy Net P h o t o s y n t h e s i s In 1983, with higher s o i l water storage, c lear sky F c was gener-a l l y higher (or more frequently pos i t i ve ) than in 1984. Comparison between 20 August 1983 and 20 August 1984 (F ig . 1.9) shows quite remarkable s i m i l a r i t y in diurnal Q, T and D, yet there was c l e a r l y higher g and F for most of 20 August 1983 (F igs . 1.2a and 1.9) . Since the only major d i f ference between these two days was in W, i t seems c lear that reduced s o i l water storage resulted in reduced F £ in August 1984. This could be due to a reduction in photosynthetic capacity or possibly to the previously noted reduction in g c causing a decrease in the supply of C0~. The work of Wong et al. (1979, - 60 -co CM o E E O CO CM CD E CO CD E o" H H H 1-•20 August 1983 20 August 1984 8 12 16 HOURS (PST) Figure 1 .9 . Comparison of net canopy photosynthesis rates (F c ) and estimated canopy i n t e r c e l l u l a r C0 2 concentration (c . ) in response to quantum f lux density (Q) and a i r temperature (T f l) for two ident i ca l c lear days over the o ld stand, with very d i f fe ren t W. During the 1983 growing season, W remained high and on 20 August was about 128 mm, whereas in 1984, W decl ined s tead i l y to about 51 mm by 20 August. The horizontal dotted l i n e p lotted at -100 yg n f 2 s " 1 on the graph of F c Indicates the prob-able contr ibut ion of s o i l respiratory C0 2 to F c . - 61 -1985a, 1985b, 1985c) shows that g g i s often correlated to photosynthetic capa-c i t y , as for example when plants are subjected to slowly imposed s o i l water d e f i c i t s . Much of the fol lowing discussion i s aimed at resolving the cause of reduced F in response to decreasing W. Farquhar and Sharkey (1982) and Jones (1985) show that the causes of changes in ass imi la t ion rate may be resolved by consideration of changes in c . . F i g . 1.9 shows that day-time c i was general ly higher, p a r t i c u l a r l y in the afternoon, on 20 August 1984 which implies that the l i m i t a t i o n to F £ in 1984 was photosynthetic rather than stomatal. Comparison of F igs . 1.8 and 1.9 shows F to be general ly s l i g h t l y lower, and day-time c. higher, on 20 August I* 1 1984 than on 29 J u l y , while temperatures were lower. This strongly suggests that the intervening four week's decrease in W was having an e f fec t on photo-synthesis . Actua l l y c. on 24 July was even higher than on 20 August 1984, ind icat ing a greater non-stomatal l i m i t a t i o n to F , but the explanation in th is case was at t r ibuted to higher T causing accumulation of f o l i a r resp i ra t -ory C 0 2 , as discussed in ( I I I .3 .1) above. Hence i t i s inferred that reduced s o i l water storage in 1984 was a factor l i m i t i n g canopy photosynthetic capa-c i t y that year . Ja rv i s and Mul l ins (1987) also concluded that reduced growth of S i tka spruce (Plcea sitohensis) on f ree ly draining s i tes in dry years was due d i r e c t l y to s o i l water d e f i c i t s rather than to ef fects of D on canopy conductance. They also point out that Watts et al. (1976) showed F £ in Si tka spruce was unaffected by D up to 1.2 kPa, because at high g s , net photosyn-thesis was l imi ted by the mesophyll conductance ( i . e . by f o l i a r photosynthetic capacity) (Beadle et al. 1981). The reasons for s o i l water storage a f fec t ing canopy photosynthetic capacity may be related to osmotic adjustment (e .g . seen c l e a r l y in Douglas-f i r seedlings by L iv ingston , 1986) since both carbohydrates and leaf proteins - 62 -may be hydrolysed to raise the osmotic potent ia l of the leaf c e l l sap. Discu-ssions of the ef fec ts of water d e f i c i t s on plant growth are presented by Hsiao et al. (1976) and Pearcy (1983). I t seems that biochemical e f fec ts of reduced leaf water potent ia l on the photosynthetic processes are not e a s i l y observed, espec ia l l y where only moderate stress occurs. However at lower p o t e n t i a l s , i t appears probable that hydrolysis of enzymes and other damage to c e l l contents are contr ibut ing f a c t o r s . Boyer (1976b) reviews many studies fol lowing Gaastra (1959) which general ly show that reduced leaf water potent ia l results in increased "mesophyll res is tance" . This i s often at t r ibuted to biochemical changes with in the ch lo rop las ts , although the precise mechanism i s s t i l l evidently undetermined. Beadle and Jarv is (1977) made In vitro measurements of electron transport rates (both photosystems I and II) and of carboxylase a c t i v i t y in extracts of needles from potted S i tka spruce seedlings subjected to water potent ia ls decreasing to as low as -3.4 MPa. The apparent lack of ef fect on In vitro e lectron transport and carboxylase a c t i v i t y was at t r ibuted mainly to the analys is technique which probably rehydrated the enzymes, demonstrating the problems associated with making v a l i d measurements of these biochemical parameters. However, they also found that leaf water potent ia ls as low as - 1 . 7 MPa were required before stomatal conductance and photosynthesis were s i g n i f i c a n t l y a f fec ted . This may seem to be in disagreement with the data presented here, but the i r resul ts were obtained at D maintained around 0.6 kPa, whereas previous work on Doug las - f i r by Tan et al. (1977) showed charac-t e r i s t i c a l l y increasing stomatal s e n s i t i v i t y to high D as Y decreased. Pr ice et al. (1986) also concluded that reduced F c of Doug las - f i r at low Y s was due, at least p a r t i a l l y , to non-stomatal causes. Moreover, i t was found when attempting to model the canopy responses (Chapter I I ) , that only r e l a -- 63 -t i v e l y weak functions of W were required to obtain best agreement between measured and predicted g and F . I t i s well establ ished that f o l i a r absc i s i c acid (ABA) concentrat-ion general ly increases in response to decreasing leaf water potent ia l (e .g . see Davies et al. 1981), which i s of considerable s ign i f i cance because ABA can evident ly exert a major inf luence over stomatal control (e .g . Raschke 1975). Furthermore, ABA i s known to d i r e c t l y inf luence photosynthetic capacity . Cornic and Miginiac (1983) have shown that stomatal and non-stomatal ef fects of ABA can be separated (using the Farquhar and Sharkey model), though both are p o t e n t i a l l y s i g n i f i c a n t in l i m i t i n g photosynthesis. The e f fec ts of low W seen c l e a r l y on 20 August 1984 (F ig . 1.9) when Y s was about - 1 . 1 5 MPa, were not apparent two weeks e a r l i e r on 6 August, when Y s « - 0 . 3 2 MPa (F ig . 1.10). However, FQ on 6 August 1984 generally exceeded that on 26 August 1983, even though Y s on the l a t t e r day was close to zero, while S was s i m i l a r for both days. Lower a i r temperature on 6 August 1984 probably explains much of the d i f fe rence , although canopy ageing d iscus -sed e a r l i e r ( I I I .1 .2) may also contr ibute . Ludlow and Jarv is (1971) found that photosynthetic capacity of S i tka spruce needles began to decl ine almost as soon as they reached matur i ty . The f indings of Wong et al. could also conceivably explain seasonal decl ines in g c evident in t h i s study as related to ( i f not caused by) the changes in canopy photosynthetic capaci ty , due both to canopy ageing and to the biochemical e f fec ts of slowly decreasing s o i l water p o t e n t i a l . - 64 -,0 4 8 12 16 20 24 HOURS (PST) Figure 1.10. Comparison of net canopy photosynthesis rates (F c) and esti -mated canopy intercellular C02 concentration (c.) in response to quan-tum flux density (Q) and air temperature (T) for two cloudy days over the old stand, with very different W. On 23 August 1983 W was about 134 mm, whereas on 6 August 1984 W was about 72 mm. The horizontal dot-- 2 - 1 ted line plotted at -100 ug m s on the graph of FQ Indicates the probable contribution of soil respiratory CO, to F . - 65 -1 1 1 . 3 . 3 E f f e c t s o f C l o u d i n e s s on Canopy Net P h o t o s y n t h e s i s Both E and F c on 6 August 1984 were greater than on 26 August 1983, even though W and Y s were lower (F igs . 1 .5 , 1 .6 , 1.7 and 1.10), so WUE was comparable, though high compared to c lear days (F ig . 1 .1) . Moreover, F c was higher throughout both these cloudy days than on any of the c lear days, and in p a r t i c u l a r , comparison of F igs . 1 .8 , 1.9 and 1.10 shows the highest rates were observed on 6 August 1984. Conditions on 6 August were evidently nearer the optimum for growth than on any other day reported here, in spi te of the ant ic ipated deleter ious e f fect of W being lower than on 26 August 1983. The factors leading to optimal condit ions would probably be lower T and greater canopy penetration by d i f fuse shortwave radiat ion on 6 August 1984 d i r e c t l y a t t r ibu tab le to the presence of cloud cover ( Jarv is et al. 1985). 1 1 1 . 3 . 4 S e a s o n a l Changes i n Canopy P h o t o s y n t h e s i s and Water Use E f f i c i ency F i g . 1.11 shows d a i l y values for F c and WUE for the measurement per iods. Since night - t ime measurements of F £ were often unre l iab le , for reasons already discussed ( I I I . 3 . 1 ) , 24-hour to ta l F £ may also be untrust -worthy. However, F i g . 1.11 shows the var ia t ion in 24-hour F c was general ly well correlated with day-time to ta l s so the 24-hour F c data are probably ser ious ly in e r ro r . I t i s therefore evident that c lear simple trends in da i l y values of F £ and WUE did not occur. However, when compared with the plot of 24 hour quantum f lux density (F ig . 1 .8) , i t i s f a i r l y apparent that the high i r radiance days of Ju ly 1984 (characterised by high T and D) had low F and consequently over 24 hours, net carbon uptake was negative. By comparison, cooler days in la te August general ly had pos i t i ve ass imi la t ion (after allowing for the contr ibut ion of R g ) , even when W was s i g n i f i c a n t l y lower. This 2 apparent inverse cor re la t ion (r « 0.4) i s a t t r ibuted to higher rates of resp i rat ion due to higher mean temperatures on days of high so lar i r rad iance . - 66 -10 20 30 16 26 5 15 25 4 AUG JUL AUG Figure 1.11. Seasonal changes In daily total net canopy photosynthesis ( F c ) , total solar Irradiance (S) and mean dally canopy water use efficiency ratio (WUE), for the young stand in early August 1983 (a), the old stand in late August 1983 (b) and the old stand during July-August 1984 (c). On the F c p lot , the sol id trace shows F £ summed for day-time hours only, whereas the dotted trace is for F£ summed over 24 hours. The horizontal broken l ine on the plot of F_ indicates the contribution of soi l resp-c _5 _ i i r a t i on assuming a mean soi l C O 2 efflux of 100 ug m s . On the plot of WUE the sol id l ine shows WUE calculated on a day-time basis (ignoring night-time respiration and evaporation) while the dotted l ine shows WUE calculated over 24 hours. - 67 -- 1 - 2 - 1 Since E was general ly about 2-3 mm d ( i . e . 2-3 kg m d ) , - 2 - 1 while F c varied around 10 g m d (on a day-time b a s i s ) , WUE was normally about +0.5% when expressed on a mass bas i s . However, some of the highest d a i l y F c and WUE occurred on p a r t i c u l a r l y overcast days. The high WUE resulted from the low E occurring on these days (due to low D), while F c was high. S imi la r resu l ts were observed for a canopy of soybeans by Baldocchi et al. (1985). Low values of WUE immediately p r io r to the high values were a t t r ibu tab le to the e f fec t of r a i n f a l l causing wet canopies, presumably the cause of a reduction in the regional ly imposed D, as can be seen from re fer -ence to F i g . 1.7. I t i s a lso apparent that F c in the old stand during late August 1983 was general ly higher than in la te August 1984. This i s in agreement with the observation reported e a r l i e r that W had s i g n i f i c a n t e f fect on F c . How-ever, day-to-day f luc tuat ions in F obscure the seasonal trend during 1984, at least over the measurement period of 6 weeks. Dai ly F c evidently increas-ed during August 1984, (presumably because the ef fects of dec l in ing W were more than compensated by reduced resp i rat ion rates due to decreasing mean d a i l y temperature). Since d a i l y E general ly decl ined during the same per iod, WUE should also have increased toward the end of August 1984. F i g . 1.11 shows that calculated WUE was general ly higher in la te August 1984, though th is trend i s not st rong, being obscured by the day-to-day f luc tuat ions in F_. - 68 -IV CONCLUSIONS It was found possible to determine canopy net photosynthesis rates (F ) over two coniferous forest stands using the modified Bowen Ratio/energy balance measurement technique. Rates of evaporation (E) and F c were general ly lower in a young stand, as compared to an older stand immediately adjacent, mainly because of d i f ferences in leaf area index. The young stand exhibited less coupling between D and LE than the old stand. However, i t a lso appears that when subjected to low D, and moder-ate to high Y s , stomatal opening in the canopy of the old stand could be substant ial enough to resul t in s i g n i f i c a n t l y greater cor re la t ion of LE to ava i lab le energy, and hence r e l a t i v e l y high values of the McNaughton-Jarvis fi parameter. Therefore in the old stand measured E was coupled c lose ly with D and r e l a t i v e l y independent of net i r radiance ( R n ) » par t l y because of the low canopy aerodynamic resistance ( r A ) , and par t l y because the calculated canopy conductance (g c ) was found to decrease in response to high D. Conversely, in the young stand g c was evidently not as responsive to changes in D, while r A was larger than in the old stand, so that E was more strongly correlated with R , and higher values of fi resu l ted . In the old stand, g was found to be only s l i g h t l y influenced by s o i l water potent ia l ( Y $ ) , though t h i s could be because very low Y g d id not occur un t i l the very end of these invest igat ions . In the old stand low Y was found to s i g n i f i c a n t l y reduce F , and from consideration of changes in the canopy e f fec t i ve sink level C0£ concen-t ra t ion ( c ^ , t h i s e f fect was independent of d i rec t e f fects of Y s on g c . Hence g was not normally a l i m i t a t i o n to F in th i s canopy. Conversely, i t i s quite possible that seasonal changes in g £ are correlated with seasonal changes in F , however caused. - 69 -While so lar i r radiance (S) was obviously an essent ia l factor gov-erning F , no simple re lat ionship between S and F was observed at the canopy l e v e l . Conversely, i t was apparent that on cloudy days F c was normally greater than on c lear days with s i m i l a r S. In so far as i t was possible to interpret night - t ime F £ measure-ments, resp i ra t ion was found to account for substant ial proportions of the day-time photosynthesis, so that 24 hour to ta l s of F c were often negative, p a r t i c u l a r l y on days characterised by high a i r temperature (T), and even a f te r correct ing for the l i k e l y contr ibut ion of s o i l respiratory CC^. Hence i t i s cer ta in that resp i ra t ion i s a major factor inf luencing the product iv i ty of the forest in general . The optimal condit ions for ,F occur on cloudy days (not heavily overcast) with low T and low D, so that a high proportion of d i f fuse shortwave radiat ion can penetrate the canopy while respiratory losses are low and canopy water use e f f i c i e n c y i s high. There were some indicat ions that g c and maximum photosynthesis rates decreased as the growing season progressed which were independent of changes in ¥ and weather. These decreases are considered to be related to changes in leaf area and physio logical changes associated with leaf ageing. - 70 -V REFERENCES Aston, A.R. 1985b. The e f fect of v e r t i c a l separation of psychrometers on the determination of Bowen Ratios over a young eucalypt f o r e s t . Agr. For. Meteor. 35: 299-307. Ayotte, K. 1984. Adjustment to step changes in f l u x . Unpubl. UBC research paper. 29 pages (includes graphs and computer program l i s t i n g ) . Baldocchi , D.D. 1987. Canopy-atmosphere water vapor exchange: can we scale from a leaf to a canopy? Paper presented at International Union of Geodesy and Geophysics Workshop on Areal Evapotranspirati on, August 1987, Vancouver, B.C. [To be publ ished] . Baldocchi , D.D., S .B . Verma and N.J . Rosenberg. 1985. Water use e f f i c iency in a soybean f i e l d : Influence of plant water s t r e s s . Ag r i c . For. Meteor. 34: 53-65. Beadle, C L . and P.G. J a r v i s . 1977. The ef fects of shoot water status on some photosynthetic p a r t i a l processes in S i tka spruce. P h y s i o l . P lant . 41: 7 -13. Beadle, C . L . , P.G. J a r v i s , H. Talbot and R.E. Nei lson. 1985a. Stomatal conductance and photosynthesis in a mature Scots pine fo res t : I I . Dependence on environmental var iables of s ingle shoots. J . Appl . Eco l . 22: 573-586. Beadle, C . L . , R.E. N ie lson, P.G. J a r v i s , and H. Talbot. 1981. Photosynthesis as re lated to xylem water potent ia l and carbon dioxide concentration in S i t k a spruce. P h y s i o l . P lant . 52: 391-400. Beadle, C . L . , H. Talbot and P.G. J a r v i s . 1982. Canopy structure and leaf area index in a mature Scots pine fo res t . Forestry 55: 105-123. Beadle, C . L . , H. Talbot , R.E. Neilson and P.G. J a r v i s . 1985b. Stomatal conductance and photosynthesis in a mature Scots pine fo res t : I I I . Var iat ion in canopy conductance and canopy photosynthesis. J . Appl . E c o l . 22: 587-595. Black, T.A. and K.G. McNaughton. 1971. Psychrometric apparatus for Bowen-ra t io determination over fo res ts . Bound. Layer Meteor. 2: 246-254. Black, T .A . , P.M. Osberg, D.T. Pr ice and D.G. G i l e s . 1986. Ef fects of sa la l removal on the water balance and tree water stress in young Douglas - f i r p lantat ions on the east coast of Vancouver Is land. MacMillan Bloedel , Woodlands Serv ices , unpublished report . Black, T .A . , D.T. P r i c e , F.M. K e l l i h e r and P.M. Osberg. 1984. Effect of understory removal on seasonal growth of a young Douglas - f i r stand. B.C.M.F. Forest Service Annual Report (1983-84). - 71 -Black, T .A . , D.T. P r i c e , P.M. Osberg and D.G. G i l e s . 1985. Effect of reduction of sa la l competition on evapotranspiration and growth of early stage Doug las - f i r p lantat ions . B.C.M.F. Forest Service Annual Report (1984-85). Boyer, J . S . 1976a. Water d e f i c i t s and photosynthesis. In T.T. Kozlowski (ed . ) . Water D e f i c i t s and Plant Growth V o l . 4. Academic Press, New York. pp. 153-190. Boyer, J . S . 1976b. Photosynthesis at low water p o t e n t i a l s . P h i l . Trans. R. Soc. Lond. B 273: 501-512. B r i x , H. 1962. The ef fect of water stress on the rates of photosynthesis and resp i rat ion in tomato plants and l o b l o l l y pine seedl ings. Phys io l . P lant . 15: 10-20. B r i x , H. 1979. Efects of plant water stress on photosynthesis and survival of four c o n i f e r s . Can. J . For. Res. 9: 160-165. Campbell, G.S. 1985. So i l Physics with BASIC, Transport models for s o i l -plant systems. E l s e v i e r , Amsterdam. 150 pages. Cornic , G. and E. Mig in iac . 1983. Nonstomatal i n h i b i t i o n of net C0 ? uptake by (+) a b s c i s i c ac id in Pharbitls nil. Plant P h y s i o l . 73: 529-533. Da Costa, J . M . N . , N.J . Rosenberg and S .B . Verma. 1986. Jo int influence of a i r temperature and s o i l moisture on CO- release by a soybean crop. A g r i c . For. Meteor. 37: 219-227. Davies, W . J . , J . A . Wilson, R.E. Sharp and 0. Osonubi. 1981. Control of stomatal behaviour in water-stressed p lants . In P.G. Jarv is and T.A. Mansfield (eds . ) . Stomatal Physiology. Cambridge Univers i ty Press, Cambridge, pp. 163-185. Denmead, O.T. 1969. Comparative micrometeorology of a wheat f i e l d and a forest of Pinvs radlata, Ag r i c . Meteor. 6: 357-371. Denmead, O.T. and E.F. Bradley. 1985. Flux gradient re lat ionships in a forest canopy. In: B.A. Hutchison and B.B. Hicks , (eds . ) . The Forest -Atmosphere In teract ion . R e i d e l , pp. 421-442. Dyer, A . J . and B.B. Hicks . 1970. Flux-gradient re lat ionships in the constant f lux layer . Quart. J . R. Met. Soc. 96: 715-721. Edwards, N.T. and P. So l i i n s . 1973. Continuous measurement of carbon dioxide evolut ion from part i t ioned forest f l o o r components. Ecology 54: 406-412. Farquhar, G.D. and T.D. Sharkey. 1982. Stomatal conductance and photosyn-t h e s i s . Ann. Rev. Plant P h y s i o l . 33: 317-345. Farquhar, G.D. and S .C . Wong. 1984. An empirical model of stomatal conduct-ance. Aust. J . Plant P h y s i o l . 11: 191-210. - 72 -F innigan, J . J . and M.R. Raupach. 1987. Modern theory of t ransfer in plant canopies in re la t ion to stomatal c h a r a c t e r i s t i c s . In E. Zeiger, G.D. Farquhar and I.R. Cowan (eds . ) . Stomatal Function. Stanford Universi ty Press, Stanford, CA. in press. Gaastra, P. 1959. Photosynthesis of crop plants as influenced by l i g h t , carbon d iox ide , temperature, and stomatal d i f f u s i o n res istance. Mededel. Landbouhogeschool Wageninen 59: 1-68. Gash, J .H .C . 1986. A note on estimating the e f fect of l imi ted fetch on micrometeorological evaporation measurements. Bound.-Layer Meteor. 35: 409-413. Gholz, H .L . , F.K. F i t z and R.H. Waring. 1976. Leaf area di f ferences asso-c iated with old-growth forest communities in the western Oregon Cascades. Can. J . For. Res. 6: 49-57 Hsiao, T .C . , E. Acevedo, E. Fereres and D.W. Henderson. 1976. Stress Metabolism. Water s t r e s s , growth and osmotic adjustment. P h i l . Trans. R. Soc. Lond. B. 273: 479-500. J a r v i s , N.J . and C E . Mull i n s . 1987. Modeling the ef fects of drought on the growth of S i tka spruce in Scotland. Forestry 60: 13-30. J a r v i s , P.G. 1985a. Transpirat ion and ass imi la t ion of tree and agr i cu l tu ra l crops: the "Omega Factor" . In M.G.R. Cannell and J . E . Jackson (eds. ) . At t r ibutes of Trees as Crop P lants . Ins t i tu te of Te r res t r ia l Ecology, Monks Wood Exp. S t n . , Abbots Ripton, U.K. J a r v i s , P.G. 1985b. Coupling of t ransp i rat ion to the atmosphere in h o r t i -cu l tu ra l crops: the Omega fac to r . Acta Hort icul turae 171: 187-205. J a r v i s , P .G . , G.B. James and J . J . Landsberg. 1976. Coniferous fo res t . In J . L . Monteith (ed . ) . Vegetation and the Atmosphere V o l . 2 . Academic Press, New York. pp. 171-200. J a r v i s , P.G. and J.W. Leverenz. 1983. Product iv i ty of temperate, deciduous and evergreen f o r e s t s . In O.L. Lange, P.S. Nobel, C B . Osmond, H. Z ieg ler (eds . ) . Physio logical Plant Ecology IV, Encyclopedia of Plant Physiology New Ser ies , Volume 12D. Spr inger -Ver lag , B e r l i n , Heidelberg, pp 233-280. J a r v i s , P.G. and K.G. McNaughton. 1986. Stomatal control of t ransp i ra t ion : sca l ing up from leaf to region. Advances in E c o l . Research 15: 1-49. J a r v i s , P .G . , R.S. Miranda and R. I . Meutze l fe ldt . 1985. Modelling canopy exchanges of water vapor and carbon dioxide in coniferous forest p lantat ions . In B.A. Hutchison and B.B. Hicks (eds . ) . The Forest-Atmosphere In teract ion . D. Reidel Publ ishing Company, pp 521-542. Jones, H.G. 1983. Plants and Microcl imate. Cambridge Univers i ty Press, Cambridge. 323 pages. Jones, H.G. 1985. P a r t i t i o n i n g stomatal and non-stomatal l im i ta t ions to photosynthesis. P lant , Ce l l and Environ. 8: 95-104. - 73 -K e l l i h e r , F .M. , T.A. Black and D.T. P r i c e . 1986. Estimating the ef fects of understory removal from a Douglas - f i r forest using a two-layer canopy evapotranspiration model. Water Resour. Res. 22: 1891-99. K l i n k a , K. 1977. Biogeocl imatic u n i t s . Nootka-Nanaimo, Sheet 92 SW, Vancouver Forest Region. B r i t i s h Columbia Min is t ry of Forests , Forest Service Research Branch, V i c t o r i a , B.C. Kirschbaum, M.U.F. and G.D. Farquhar. 1984. Temperature dependence of whole-leaf photosynthesis in Eucalyptus pauolflora S ieb . ex Spreng. Aust. J . Plant P h y s i o l . 11: 519-538. Kozlowski , T.T. 1982. Water supply and tree growth. Part I water d e f i c i t s . Review A r t i c l e , For. Abs. 43: 57-95. Leuning, R., O.T. Denmead, A .R.G. Lang and E.H. Ohtaki . 1982. Effects of heat and water vapor transport on eddy covariance measurement of C0 ? f l uxes . Bound.-Layer Meteor. 23: 209-222. Leverenz, J.W. 1981. Photosynthesis and t ransp i rat ion in large Doug las - f i r : diurnal v a r i a t i o n . Can. J . Bot. 59: 349-356. Leverenz, J .W., J .D. Deans, E.D. Ford, P.G. J a r v i s , R. Milne and D. Whitehead. 1982. Systematic spat ia l va r ia t ion of stomatal conductance in a S i tka spruce p lan ta t ion . J . Appl . E c o l . 19: 835-851. L indroth, A. 1984. Seasonal var ia t ion in pine forest evaporation and canopy conductance. Ph.D t h e s i s . Univers i ty of Uppsala, Uppsala, Sweden. L indroth , A. 1986. Influence of var iat ions in needle area on pine forest water balance. A g r i c . For. Meteor. Unpublished d r a f t . 23 pages i n c l u -ding graphs. L iv ingston , N .J .L . 1986. Water r e l a t i o n s , surv ival and growth of conifer seedlings planted on a high elevat ion south-facing c learcut . Unpubl. Ph.D. t h e s i s . Univers i ty of B r i t i s h Columbia, Vancouver, B.C. Ludlow, M.M. and P.G. J a r v i s . 1971. Photosynthesis in S i tka spruce (Picea s i tchens is (Bong.) C a r r . ) . I. General c h a r a c t e r i s t i c s . J . Appl . E c o l . 8 : 925-953. McCaughey, J . H . 1985. Energy balance storage terms in a mature mixed forest at Petawawa, Ontario. A case study. Bound.-Layer Meteor. 31: 89-101. Mcintosh, D.H. and A . S . Thorn. 1969. Essent ia ls of Meteorology. Wykeham, London. 239 pages. McMurtrie, R . E . , S . L inder , M.L. Benson and L. Wolf. 1986. A model of leaf area development for pine stands. In T. Fujimori and D. Whitehead (eds . ) . Proceedings of Crown and Canopy Structure in Relation to Product iv i ty Symposium, 1986. Forestry and Forest Products Research I n s t i t u t e , Ibarak i , Japan, pp. 284-307. - 74 -McNaughton, K.G. and T.A. Black. 1973. A study of evapotranspiration from a Doug las - f i r forest using the energy balance approach. Water Resour. Res. 9: 1579-1590. McNaughton, K.G. and P.G. J a r v i s . 1983. Predict ing e f fec ts of vegetation changes on t ransp i rat ion and evaporation. In T.T. Kozlowski (ed. ) . Water D e f i c i t s and Plant Growth V o l . 7. Academic Press, New York. pp. 1-47. Meek, D.W., J . L . H a t f i e l d , T.A. Howell, S .B . Idso and R . J . Reginato. 1984. A general ized re lat ionsh ip between photosynthet ical ly act ive radiat ion and solar r a d i a t i o n . 76: 939-945. Mi lne , R., J . D . Deans, E.D. Ford, P.G. J a r v i s , J . Leverenz and D. Whitehead. 1985. A comparison of two methods of estimating t ransp i rat ion rates from a S i tka spruce p lantat ion . Bound.-Layer Meteor. 32: 155-175. Miranda, A . C . , P.G. Ja rv i s and J . Grace. 1984. Transpirat ion and evaporation from heather moorland. Bound.-Layer Meteor. 28: 227-243. Monteith, J . L . 1965. Evaporation and environment. In G.E. Fogg (ed . ) , The State and Movement of Water in L iv ing Organisms. 19th Symposium of the Society for Experimental Biology. Cambridge Univers i ty Press, London, pp. 205-235. Monteith, J . L . 1973. P r inc ip les of Environmental Physics . Edward Arnold, London. 241 pages. Monteith, J . L , G. Szeicz and K. Yabuki. 1964. Crop photosynthesis and the f lux of carbon dioxide below the canopy. J . Appl . E c o l . 1: 321-337. Ninomiya, I. and K. Hozumi. 1981. Respirat ion of forest trees (I) Measure-ment of resp i rat ion in Plnus densi-thunbergil UYEKI by an enclosed standing tree method. J . Jap. For. Soc. 63: 8 -18. Osberg, P.M. 1986. Lysimeter measurements of sa la l understory evapotrans-pi rat ion and forest s o i l evaporation a f te r sa la l removal in a Douglas-f i r p lan ta t ion . Unpubl. M.Sc. t h e s i s , Univers i ty of B r i t i s h Columbia, Department of So i l Science, Vancouver, B.C. Parkinson, K . J . and B . J . Legg. 1978. Ca l ib ra t ion of i n f r a - r e d analysers for carbon d iox ide . Photosynthetica 12: 65-67. Pearcy, R. W. 1983. Physio logical consequences of c e l l u l a r water d e f i c i t s : nonstomatal i n h i b i t i o n of photosynthesis by water s t r e s s . In H.M. Taylor , W.R. Jordan and T.R. S i n c l a i r (eds . ) . L imitat ions to e f f i c i e n t water use in crop production. American Society of Agronomy, Madison, WI. P r i c e , D.T., T.A. Black and F.M. K e l l i h e r . 1986. Effects of sa la l understory removal on photosynthetic rate and stomatal conductance of young Doug las - f i r t rees . Can. J . For. Res. 16: 90-97. - 75 -P r i e s t l e y , C.H.B. and R . J . Taylor . 1972. On the assessment of surface heat f l u x and evaporation using la rge -sca le parameters. Monthly Weather Review 100: 81-92. Raschke, K. 1975. Simultaneous requirement of carbon dioxide and absc is i c ac id for stomatal c los ing in Xanthium strvmarivm L. Planta 125: 243-259. Raupach, M.R. and B . J . Legg. 1984. The uses and l i m i t a t i o n s of f lux -gradient re la t ionsh ips in micrometeorology. Ag r i c . Water Mgmt. 8: 119-131. Schulze, E. -D, M.I. Fuchs and M. Fuchs. 1977. Spat ia l d i s t r i b u t i o n of photo-synthet ic capacity and performance in a mountain spruce forest of northern Germany. Oecologia (Ber l . ) 29: 43-61. Spi t t lehouse, D.L. and T.A. Black. 1979. Determination of forest evapotrans-p i r a t i o n using Bowen ra t io and eddy cor re la t ion measurements. J . Appl . Meteor. 18: 647-653. Spi t t lehouse, D.L. and T.A. Black. 1980. Evaluation of the Bowen rat io/ energy balance method for determining forest evapotranspirat ion. Atmosphere-Ocean 18: 98-116. Stewart, J . B . and A . S . Thorn. 1973. Energy budgets in pine f o r e s t . Quart. J . R. Met. Soc. 99: 134-170. Tan, C S . and T.A. Black. 1976. Factors a f fec t ing the canopy resistance of a Douglas - f i r f o r e s t . Bound.-Layer Meteor. 10: 475-488. Tan, C . S . , T.A. Black and J . U . Nnyamah. 1977. Character i s t i cs of stomatal d i f f u s i o n resistance in a Douglas f i r forest exposed to s o i l water d e f i c i t s . Can. J . For. Res. 7: 595-604. Tan, C . S . , T.A. Black and J . U . Nnyamah. 1978. A simple d i f f u s i o n model of t ransp i rat ion applied to a thinned Doug las - f i r stand. Ecology 59: 1221-1229. Tang, P .A . , K.G. McNaughton and T.A. Black. 1974. Inexpensive diode thermo-metry using integrated c i r c u i t components. Can. J . For. Res. 4: 250-254. Tanner, C.B. 1981. Transpirat ion e f f i c i e n c y of potato. Agro. J . 73: 59-64. Thorn, A . S . 1975. Momentum, mass and heat exchange of plant communities. In J . L . Monteith (ed . ) . Vegetation and the Atmosphere. V o l . 1. P r i n c i p l e s . Academic Press, London, pp. 57-109. Thorn, A . S . , J . B . Stewart, H.R. O l i ver and J .H .C . Gash. 1975. Comparison of aerodynamic and energy budget estimates of f luxes over a pine fo res t . Quart. J . R. Met. Soc. 101: 93-105. Troeng, E. and S. L inder . 1982b. Gas exchange in a 20-year -o ld stand of Scots pine: I I . Var iat ion in net photosynthesis and t ranspi rat ion with in and between t rees . P h y s i o l . P lant . 54: 15-23. - 76 -Verma, S . B . , D.D. Baldocchi , D.E. Anderson, D.R. Matt and R . J . Clement. 1986. Eddy f luxes of C 0 2 , water vapor and sensible heat over a deciduous f o r e s t . Bound.-LaVer Meteor. 36: 71-91. Waring, R.H. and J . F . F r a n k l i n . 1979. Evergreen coniferous forests of the P a c i f i c Northwest. Science 204: 1380-1386. Watts, W.R., R.E. Neilson and P.G. J a r v i s . 1976. Photosynthesis in S i tka spruce (Picea s i tchens is (Bong.) c a r r . ) : VI I . Measurements of stomatal conductance and 14C02 uptake in a forest canopy. J . Appl . E c o l . 13: 623-638. Webb, E.K. 1970. P r o f i l e re la t ionsh ips : the l o g - l i n e a r range and extension to strong s t a b i l i t y . Quart. J . R. Met. Soc. 96: 67-90. Webb, E .K . , G . I . Pearman and R. Leuning. 1980. Correction of f lux measure-ments for density e f fects due to heat and water vapour t rans fe r . Quart. J . R. Met. Soc. 106: 85-100. Whitehead, D., P.G. Ja rv is and R.H. Waring. 1984. Stomatal conductance, t r a n s p i r a t i o n , and resistance to water uptake in a Plnvs sylvestrls spacing experiment. Can. J . For. Res. 14: 692-700. Wong, S . C , I.R. Cowan and G.D. Farquhar. 1979. Stomatal conductance corre-lates with photosynthetic capacity . Nature 282: 424-426. Wong, S . C , I.R. Cowan and G.D. Farquhar. 1985. Leaf conductance in re lat ion to rate of C0 2 a s s i m i l a t i o n . I. Influence of nitrogen n u t r i t i o n , phos-phorus n u t r i t i o n , photon f lux densi ty , and ambient p a r t i a l pressure of C0 2 during ontogeny. Plant P h y s i o l . 78: 821-825. Wong, S . C , I.R. Cowan and G.D. Farquhar. 1985. Leaf conductance in re lat ion to rate of C0 2 a s s i m i l a t i o n . I I . E f fects of short-term exposures to d i f f e r e n t photon f lux d e n s i t i e s . Plant P h y s i o l . 78: 826-829. Wong, S . C , I.R. Cowan and G.D. Farquhar. 1985. Leaf conductance in re la t ion to rate of C0 2 a s s i m i l a t i o n . I I I . Influences of water stress and photo inh ib i t ion . Plant P h y s i o l . 78: 830-834. Yoda, K. 1967. Comparative ecological studies on three main types of forest vegetation in Thai land. I I I . Community r e s p i r a t i o n . Nature and L i fe in Southeast Asia 5: 83-148. Yoda, K. 1971. Forest Ecology. Tsuk i j i - shokan, Tokyo. 331 pages. [In Japanese o n l y ] . Yoda, K. 1978. Estimation of community r e s p i r a t i o n . In T. K i r a , Y. Ono and T. Hosokawa (eds . ) , JIPB Synthesis V o l . 18. B io log ica l production in a warm temperate evergreen oak forest of Japan. - 77 -CHAPTER_II DEyELOPMENT_AND_EV EyAPORATION_AND_PHOIOSYN I_INIRODUCIION Rel iab le predict ion of forest product iv i ty for economic and mana-ger ia l purposes i s a problem occupying many researchers. Most methods in common use re ly on y i e l d tables or models derived from mensurational data obtained from forest sample p lots of d i f fe rent ages growing on s i tes of vary-ing p roduct i v i t y . In North America, normal y i e l d tables ( i . e . for natural f u l l y stocked stands) were f i r s t developed for Doug las - f i r as ear ly as 1930 (McArdle et al. 1961), and have since been revised and improved extensively (Bare 1986). The major l i m i t a t i o n of these tables i s that because they are derived from natural stand data , they are probably not su i tab le for predict ing the product iv i ty of second growth p lantat ions . By comparison the U.K. Forestry Commission has developed managed y i e l d tables for a l l species of economic importance in B r i t a i n , using data obtained from y i e l d p lots set up in plantat ions dating back to the ear ly 1920s (Hamilton and C h r i s t i e 1971). Recently, a range of computer growth and y i e l d models has been developed in the P a c i f i c Northwest, many s p e c i f i c a l l y intended to predict y i e l d of managed Doug las - f i r p lantat ions . Prominent examples are DFSIM ("Douglas-Fir Simulator" (Curt is et al. 1981)), TASS ("Tree And Stand Simulator" (Mitchel l and Cameron 1985; Mi tche l l 1986)) and FORCYTE ("Forest Nutr ient Cycl ing and Y ie ld Trend Evaluator" (Kimmins et al. 1986)). While - 78 -there are many s i g n i f i c a n t di f ferences in the approaches used in each of these and other models, they have a common basis in that they are derived d i r e c t l y from and/or ca l ib rated against biometric data obtained from y i e l d studies. Such models general ly simulate growth processes with l i t t l e attent ion to physiology, and c h a r a c t e r i s t i c a l l y have output in terva ls of the order of one year. Environmental influences are considered only as an amalgamation of loca l c l i m a t i c and edaphic conditions manifested in the " s i t e q u a l i t y " . The end resu l t of such models i s a family of y i e l d curves that attempts to give a best estimate of standing biomass and/or timber volume at d i f fe ren t ages in the l i f e of the stand, possibly influenced by management p resc r ip t ions . While there i s nothing wrong with t h i s approach from a p rac t i ca l viewpoint, i t cannot account very well for short-term changes in s i t e condit ions (of the order of one day, e . g . changes in weather cond i t ions ) , and such models may be quite inappropriate for use on s i t e s and species not investigated in the o r ig ina l y i e l d s tud ies . The a l te rnat i ve approach i s to t ry to understand in deta i l the ef fects of a l l topographic, edaphic and c l imat i c var iables (const i tut ing "the s i t e " ) on the physio logical processes of dominant species in the forest eco-system, so that product iv i ty can then be estimated from measured s i t e data and species c h a r a c t e r i s t i c s . A few years ago, t h i s approach was barely conceiv-ab le . The complexity of the approach was such that even with r e l a t i v e l y homogeneous a g r i c u l t u r a l systems a working process-or iented model took many years of e f f o r t to perfect (e .g . see Lemon (1983) for a review). However recent and continuing developments are such that a phys io logical ly -based forest growth model i s becoming a real p o s s i b i l i t y , which might be developed into a version su i tab le for everyday managerial use. - 79 -The term "empi r ica l " has been avoided so f a r , because as many researchers have ind icated , a l l models contain empirical elements ( i . e . r e l a -t ionships derived from observations or measurements without any reference to the underlying mechanism) (Landsberg 1981, 1986)). Both Landsberg and Charles-Edwards et al. (1986) allow that i t i s feas ib le to make sat is factory long-term predict ions of the product iv i ty of a crop or forest without under-standing the contr ibut ing processes. However i t i s only through more deta i led analys is and simulation at lower leve ls of complexity that a f u l l explanation of longer term responses can be achieved. This should also al low more suc-cessful extension of a high level model (e .g . of a whole stand over a complete forest rotat ion) to other species and other s i t e condi t ions . A review shows that such models are current ly under development, with many forest ecosystem processes being modelled at low leve ls of comp-l e x i t y , so that combining the best of these d i f fe ren t submodels should u l t i -mately be poss ib le . Solar i r radiance (S) a r r i v i n g at the top of the canopy can be predicted accurately for any time at any s i t e (Campbell 1977, 1985; Gates 1980) and there are several semi-empirical methods of predict ing the proportions of d i f fuse and d i rec t radiat ion (Spi t ters 1986; Bristow and Campbell 1985; Gates 1965, 1980; Liu and Jordan 1960). Recent and ongoing work by Norman and coworkers (Norman and Ja rv i s 1975; Lang et al. 1985) has provided a sound theoret ica l basis for estimating how l i g h t penetrates a vegetation canopy, while work by Leverenz and Jarv is (1979, 1980), and by Oker-Blom and coworkers (Oker-Blom 1985; Oker-Blom et al. 1983; Smolander et al. 1987) should enable accurate predict ions of how photosynthesis varies with the or ientat ion of indiv idual needles. Models have also been developed to enable predict ion of seasonal changes in canopy conductance (Lindroth 1986) - 80 -and canopy photosynthesis (McMurtrie e t a l . 1984) as a function of changes in leaf area. The physiology of photosynthesis at the c e l l u l a r level i s r e l a -t i v e l y well understood, and recent work by Farquhar and coworkers (e .g . Farquhar and Sharkey 1982; Caemmerer and Farquhar 1981; Farquhar and Caemmerer 1982; Kirschbaum and Farquhar 1984; Brooks and Farquhar 1985) i s rapidly gaining wide acceptance as being the most correct analys is to date. In p a r t i c u l a r , Wong e t a l . (1979, 1985a, 1985b, 1985c) have shown c lea r l y that under many condi t ions , stomatal functioning i s remarkably correlated to f o l i a r photosynthesis ra te , although exceptions have been observed by Kuppers and Schulze (1985). Farquhar and Wong (1984) have therefore proposed an "empi r ica l " model where stomatal conductance i s large ly contro l led by the photosynthesis rate . This work implies that the commonly held view that stomata normally l i m i t photosynthesis i s probably incor rec t , and furthermore, that f o l i a r t ransp i rat ion may even be governed by the photosynthesis rate . Recent re-examination of the evidence surrounding crop water re lat ions has led Monteith (1986) and others (Cowan 1982) to propose s i m i l a r hypotheses for whole canopies. However, there i s no doubt that under f i e l d condi t ions , diurnal var ia t ion in atmospheric vapour pressure d e f i c i t can have s i g n i f i c a n t ef fects on stomatal behaviour, at least for coniferous forest (Beadle e t a l . 1985; Whitehead e t a l . 1984; Ja rv is 1985; Watts e t a l . 1976; Tan and Black 1976; Grace e t a l . 1975), which suggests that the re lat ionsh ip between photosyn-thesis and canopy conductance may not be stra ightforward. Pr ice e t a l . (1986; Appendix X) attempted to show that the physio logical models of Caemmerer and Farquhar (1981) and Farquhar and Sharkey (1982) could be used with some success to predict leaf photosynthesis rates of Douglas - f i r trees in - 81 -the f i e l d , and concluded that conductance was un l ike l y to be l i m i t i n g to photosynthesis under most condit ions. Very few canopy-level photosynthesis models published to date ( i f any) have attempted to use these physiological models as d r i v ing funct ions . Chapter I reported on the resu l ts of measurements of forest canopy evaporation and net photosynthesis covering periods of two and s i x weeks in two consecutive growing seasons. Canopy conductances were a lso estimated from the evaporation rates using the Penman-Monteith combination equation (Monteith 1965). The overal l object ive of th i s chapter i s to demonstrate that a physio-log ica l approach can be used to model canopy net photosynthesis successfu l l y . To achieve t h i s aim, i t would also be very desi rable to model canopy conduct-ance from physio logical p r i n c i p l e s , but since no p rac t i ca l mechanistic model of stomatal funct ioning ex is ts yet (Farquhar and Wong 1984; Ja rv i s e t a l . 1985), the conductance submodel developed and used here i s empirical in nature. The dr i v ing var iables are considered to be meteorological , measured on a ha l f -hour l y basis and s o i l water storage, monitored d a i l y . The l a t t e r could be calculated using a water balance model, such as Spitt lehouse and Black (1981) or K e l l i h e r e t a l . (1986). Testing of the model w i l l use the s o i l water storage and weather data reported in Chapter I. S p e c i f i c a l l y , the objectives of t h i s chapter are to : (1) describe the canopy conductance and net photosynthesis submodels, and a simple canopy radiat ion submodel, (2) evaluate the conductance submodel for predict ing canopy conductance and evaporation, and (3) evaluate the canopy net photosyn-thesis submodel for predict ing C0 7 exchange and water use e f f i c i e n c y . - 82 -II_STRyCTyRE_OF_THE_MODEL The purpose of the model was to create a simple representation of canopy processes to enable predict ion of diurnal evaporation and photosyn-thes is at the stand canopy l e v e l , i . e . of generating reasonably accurate h a l f -hourly estimates of the f lux dens i t ies of water vapour and C0 2 at the top of the canopy. The model consists of four submodels: (1) a so lar radiat ion sub-model used to a l loca te the measured solar i r rad iance , S , into separate d i rect and d i f fuse components at the top of the canopy, S b and S d respect ive ly ; (2) a canopy layer quantum f lux density submodel used to attenuate the photosyn-t h e t i c a l l y act ive wavelengths of S b and S d with increasing depth into the canopy; (3) a leaf conductance submodel operating at each canopy layer , which predicts stomatal conductance from the modelled quantum f lux density and other measured v a r i a b l e s ; (4) a leaf photosynthesis submodel derived from the work of Farquhar and Caemmerer (1981) and others, a lso operating at each canopy layer . The model d i f f e r s s i g n i f i c a n t l y from that described by Jarv is et a l . (1985), since the l a t t e r does not employ the same physio logical basis for modelling photosynthesis, while the i r canopy l i g h t penetration submodel i s considerably more sophist icated than the one presented here. A complete source l i s t i n g of the current model program i s given in Appendix VI I I . - 83 -S o l a r_I L rradj_ance_Submo For each ha l f -hour l y time step, measured S i s passed to the rout-ine , together with the Ju l ian Day, local standard t ime, la t i tude and longitude and the local barometric pressure, P a . The solar e levat ion angle, 8 , i s a determined using standard equations (e .g . L i s t 1971; Campbell 1977, 1985; Gates 1980), and used to ca lcu late the solar i r radiance received at the outer edge of the atmosphere (para l le l to the planetary sur face) , S t t , from _2 1360 s ine , where 1360 i s the assumed solar constant in W m . Assuming a c lear sky, S t t was separated into d i rec t and d i f fuse components at the surface, S b t and S d t respect ive ly , according to empirical equations of Liu and Jordan (1960) presented in Gates (1965 and 1980): S b t = S t t ^ CD S d t = S t t ( 0 , 2 7 1 " O - 2 9 4 1 " 1 ) (2) where i i s the atmospheric transmission c o e f f i c i e n t , estimated from the data to be 0.725 1 , and m i s the opt ica l a i r mass number, calculated from (Gates 1980; Campbell 1977): m = P a / (P S sine) (3) where P g i s standard barometric pressure (101.3 kPa). I f 6 should happen to be very close to zero a problem can occur because l/s in6 tends to i n f i n i t y , but t h i s may be avoided conveniently by assuming that a mountainous skyl ine w i l l ra ise the elevat ion angle at sunr ise . Accordingly sunrise was assumed to occur when 6 reached 2°. The approximate value of T was subsequently confirmed from work by Hay (1986, personal communi-cation) who observed that eruption of the Mexican volcano El Ch1ch6n during 1981-82, caused meas-urable reductions 1n atmospheric transm1ssiv1ty In the Vancouver area during 1983 and 1984. - 8 4 -To al low f o r the ef fect of cloud cover on the proportions of d i rec t and d i f fuse r a d i a t i o n , the values of S and S T T were compared. If S < 0 . 1 S T T , then the sky was considered to be f u l l y overcast, so S D was set equal to S . Conversely, i f S > 0 . 7 5 S T T then the sky was assumed to be completely c lear and S D calculated as s ( s d t / ( s d t + s b t ) ) . For S in the range 0 . 1 S J . J . i S & 0 . 7 5 S ^ , S D was estimated using an empirical re lat ionsh ip derived from the work of L iu and Jordan ( 1 9 6 0 ) . An exponential f i t was used on the i r graphical r e l a t i o n s h i p , with resul ts as fo l lows : S D = 1 . 3 6 8 1 6 S exp [ - 2 . 6 9 8 3 5 4 S / S T T ] ( 4 ) 2 which y ie lded r of 0 . 9 9 1 3 with 8 points on the graph. F i n a l l y , S B was calculated as ( S - S D ) . On a ha l f -hour l y b a s i s , the separation between S B and S D can only be approximate, but i t i s considered essent ia l that an attempt be made in view of the s ign i f i cance of d i f fuse shortwave radiat ion penetration from the top of the canopy for photosynthesis at lower leve ls ( Jarv is et al. 1 9 8 5 ) . An a l te rna t i ve equation for the separation of d i rec t and d i f fuse shortwave radiat ion was also published by Bristow and Campbell ( 1 9 8 5 ) , which could be used in t h i s radiat ion model. However, since the equation i s also intended for use on d a i l y i r r a d i a t i o n t o t a l s , i t i s un l ike l y to provide demonstrably better resul ts than the approach based on L iu and Jordan's data. The outputs from th is submodel are S D and S B at the point of pene-t ra t ion of the topmost layer of f o l i a g e . II*_2 C a n o p . y _ _ L e y e l _ Q u a n t u The inputs to t h i s submodel are S B and S D at the top of the canopy and the cumulative leaf area index, L a „ , determined as the sum of leaf area ac index (LJ increments from the top of the canopy. For each L, increment, a a - 85 -assuming random fo l iage d i s t r i b u t i o n , the mean leaf level quantum f lux densi ty , Q, i s estimated from (Jones 1983): Q = P b ( S b e x p [ - K b L a c ] ) + P d ( S d e x p [ - K d L a c ] ) (5) where K b and < d are the canopy ext inct ion c o e f f i c i e n t s for d i rec t and d i f fuse shortwave radiat ion and P b and P d are the assumed quantum energy contents of the d i rec t and d i f fuse components respect ive ly . From consideration of data for a s i m i l a r Doug las - f i r canopy at Courtenay, B .C . , presented by Hardy (1975) and values reported in Ja rv is e t a l . (1976), average K b and K d were estimated at 0.58 and 0.55 respect ive ly , while P b and P d were taken to be 1.90 and - 2 - 1 - 1 2.20 umol m s W respect ive ly , a f te r Meek e t a l . (1984). The value for K d includes an ext inct ion c o e f f i c i e n t of 0.4 for depletion of photosynthetic-a l l y ac t i ve rad iat ion (PAR) by the canopy, estimated from data in Jarv is et a l . (1976). llj.3 L e a f - l e v e l _ S torn a t a l _ C o ^ This submodel i s e n t i r e l y empirical in nature. The measured canopy conductances for water vapour were determined from energy balance data, using the Penman-Monteith equation (see Chapter I) while canopy aerodynamic conductance (g A) was estimated from the wind ve loc i t y using the procedure described at Appendix I I I . The submodel estimates canopy conductance from measured values of physical environmental va r iab les , consistent with stomatal conductance c h a r a c t e r i s t i c s determined by Tan et a l . (1977) and by K e l l i h e r (1985) for a s i m i l a r stand of Douglas - f i r at Courtenay, B.C. A l l ha l f -hour l y measured canopy conductance, g c > data points (af ter e l iminat ion of values considered unre l iable) were p lotted against measured values of the var iables considered to be important factors a f fect ing stomatal conductance. These were root-zone s o i l water storage, W, atmospheric - 86 -vapour pressure d e f i c i t , D, and solar i r rad iance , S. (Note however, that quantum f lux density above the canopy, Q, was calculated from S, and used as the i r radiance input var iable for both the canopy conductance and photosyn-thesis submodels). So i l water storage was treated as a d r i v ing var iable in preference to s o i l water p o t e n t i a l , Y s , because W was derived d i r e c t l y from neutron hydroprobe measurements (assuming a root-zone depth of 820 mm), where-as ¥ s was merely estimated from a combination of laboratory-determined reten-t ion curves and incomplete s o i l psychrometric data (see Chapter I ) . Neverthe-l e s s , i t would be convenient to use Y s as a dr i v ing var iable i f required. Time since dawn, t , apparently influenced stomatal conductance in mid to la te afternoon, which was also observed by Liv ingston (1986), although the reasons for t h i s are not c l e a r . L iv ingston suggested that day-time a b s c i s i c ac id (ABA) accumulation by leaf guard c e l l s might have t h i s e f f e c t , re fer r ing to work of Wright (1977) and Sivakumaran et a l . (1980), who have shown that endogenous ABA concentrations increase s i g n i f i c a n t l y in water-stressed p lants . Hsiao (1976) showed that ABA has an important role in guard c e l l ion transport and that high concentrations are correlated with stomatal closure (hence the frequent research use of ABA as an a n t i t r a n s p i r a n t ) . Another possible reason for the e f fect of time i s the loca l i sed depletion of water in the rhizosphere causing diurnal va r ia t ion in rhizosphere s o i l water p o t e n t i a l , which would be expected to reach a minimum during the afternoon. Redist r ibut ion of s o i l moisture late in the day and overnight ( l imi ted by low hydraul ic conduct iv i ty in dry s o i l s ) would resul t in the conductance recovering to i t s ear ly morning maximum the next day. The exact mechanism for such a hypothesis i s unclear. K e l l i h e r e t a l . (1984), working on young Doug las - f i r trees at Courtenay, found that day-time evaporation caused s i g n i f i c a n t decreases in leaf water p o t e n t i a l , which resulted in a poor - 87 -cor re la t ion between leaf water potent ial and g c , although the l a t t e r was c l e a r l y dependent on Y s . Recent elegant work by Gollan e t a l . (1986) has shown that stomatal conductance can respond d i r e c t l y to low s o i l water content even when water potent ia l gradients across the plant have been a r t i f i c i a l l y e l iminated, which he suggests may indicate that the roots can communicate d i r e c t l y with the stomata by a chemical messenger system. For each v a r i a b l e , a boundary l i n e function ( J a r v i s , 1976; Jones 1983) was f i t t e d to the data , where points f a l l i n g below the l i n e were regard-ed as being l imi ted by some combination of other f a c t o r s . The plots of g c against D, W and S are shown in F i g . 2 . 1 a - c , together with the boundary l ine functions f i n a l l y used in the model, referred to as g ^ D ) , g2(W) and g 3 (S ) . While the boundary l ines are r e l a t i v e l y c lear for g^D) and g 2 ( S ) , F i g . 2.1b does not suggest a strong dependence on W, at least for the range of W where most measurements were made. However when the data of F i g . 2.1b were s t r a t i -f i ed for d i f fe ren t ranges of D, the dependence of g c on W became much c learer , and a ser ies of curves could be f i t t e d . The boundary l i n e function presented in F i g . 2.1b was derived from careful f i t t i n g of a s ingle function to th is ser ies of curves. The most su i table function to be used for g 4 ( t ) was deter -mined to be g„ = g___,.(l - 0.066t) where t i s the time in hours since sunrise C C i l i a X and g c m a x i s the assumed maximum bulk (stomatal) conductance of the dry canopy. At night g 3 (S) always returned zero, so g 4 ( t ) was a r b i t r a r i l y set to g^m a v' Following Ja rv i s (1976), the modelled canopy conductance could then be estimated as : 9cmod " 9cmax 9 1 (D)g 2 (W)g 3 (S)g 4 ( t ) (6a) - 88 -Figure 2 . 1 a . Var ia t ion of canopy conductance (g c ) with atmospheric vapour pressure d e f i c i t (D) for the old stand at Dunsmuir Creek during 1983 and 1984. A l l the p lotted points represent s ing le ha l f -hour l y values es t -imated from the measured evaporation rates using the Penman-Monteith equation. The s o l i d l i n e represents the "boundary l i n e " given by 9cmod = 9cmax 2 , 2 4 3 4 exp[ -1 .6D 0 " 4 5 ] as used in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was s m a l l . - 89 -25 0 40 80 120 160 Soil water storage (mm) Figure 2.1b. Var iat ion of canopy conductance (g c ) with s o i l water storage (W) for the old stand at Dunsmuir Creek during 1983 and 1984. A l l the p lot ted points represent s ingle ha l f -hour l y values. The s o l i d l ine represents the "boundary l i n e " function g . = g „ „ , 0.3344W 0" 2 3 as used acmod 3cmax in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was s m a l l . - 90 -Figure 2 .1c . Var iat ion of canopy conductance (g c ) with so lar i r radiance (S) fo r the old stand at Dunsmuir Creek during 1983 and 1984. A l l the p lot ted points represent s ingle ha l f -hour l y values. The s o l i d l ine represents the "boundary l i n e " function g . = g ^ , „ 2 . 0 4 S " 0 " 0 0 3 as used J 'cmod 3cmax in the conductance submodel. Shaded points occurred when the canopy was wet, or when ava i lab le energy was smal l . Note that for the model, photon f l u x density (Q) was used as a d r i v ing var iab le even though the funct ion shown here re lates g„ to S . The constant 2.04 represents the c _ i average photon content of above-canopy so lar i r radiance in umol W . - 91 -During the analys is (F ig . 2.1a) i t was observed that the depend-ence of g on D exhibited some hysteres is , with g often being lower in the afternoon than in the morning at an equivalent value of D. A i r temperature, T, was therefore considered as another possible factor inf luencing conductance independently of D. Af ter determining a re lat ionship between gQ and T, of the form: 9c - 9 c m a x ( e x p [ e 2 ( T g m a x - T)]) ; T < T g m a x (7a) 9 C = 9 c m a x d - e ^ T ^ - T) 2 ) ; T > T g m a x (7b) where g „ m a v occurs at temperature T „ m . (about 17.0 °C), i t was concluded Liiiax gmax that the observed hysteresis was due to the vapour pressure, e, dec l in ing during afternoon (presumably as regional evaporation decreased while the height of the mixed layer increased) so that the equivalent value of D occurred at a lower temperature below T g m a x . However, including a function of t h i s form did not resul t in any s i g n i f i c a n t improvement in the cor re lat ion between modelled and measured data a f te r inc lus ion of g 4 ( t ) . The boundary l i ne technique of (6a) was used successful ly for analys is of Doug las - f i r seedling data at Cameron Lake by Livingston (1986). However, in the present study i t was found unsat is factory , because some meteorological var iables were often highly correlated and the i r influences on conductance s i m i l a r ( p a r t i c u l a r l y D and T discussed above). Hence i t was often the case that when two or more var iables were l i m i t i n g , the product of the separate functions would s i g n i f i c a n t l y underestimate the measured value, as discussed at length by Ja rv i s (1976). An a l te rnat i ve approach was used, where instead of ca lcu la t ing the product of the funct ions , the s ingle function y i e l d i n g the lowest value was selected. Thus: 9cmod = 9 c m a x [min imum{g 1 (D) ,g 2 (W) ,g 3 (S) ,g 4 ( t ) } ] (6b) - 92 -This "minimum funct ion" approach has the advantage of being even simpler than (6a). Since only a s ingle function i s used in any par t i cu la r predict ion of g c , i t assumes there i s no in teract ion between physical factors which further reduces the conductance. This approach, sometimes referred to as " l i m i t i n g factor a n a l y s i s " , i s e s s e n t i a l l y Blackman's (1905) p r inc ip le of " l i m i t i n g fac tors" appl ied to stomatal conductance, which might be reasonable in view of the recent work of Wong et al. (1979, 1985a, 1985b, 1985c) who found strong cor re lat ions between stomatal conductance and photosynthesis rates for a wide range of higher p lants , and considering that photosynthesis rate i s i t s e l f often l imi ted to the minimum rate of several interdependent physio logical processes. I t should be noted that some " f ine - tun ing" of the c o e f f i c i e n t s was found necessary once the models were working, in order to optimise the f i t between measured and modelled data. This appl ied equally to (6a) and (6b), although the c o e f f i c i e n t s f i n a l l y selected (reported in F i g . 2.1 and above) were optimised for (6b) since th is produced better r e s u l t s . Since the canopy was divided into n layers corresponding to 1 m leaf area increments, modelled g c was obtained by assuming that the boundary l i n e functions in (6) were equally appl icable to a l l n layers . Thus the ef fects of D, W, and t were assumed constant for a l l layers , while the canopy l i g h t penetration model ( I I .2) was applied to estimate the mean layer photon f lux dens i ty , Q. The boundary l i n e function from F i g . 2.1c was used to determine g 3(Q) (assuming Q = 2.04S). Mean layer g $ was then determined using (6) with g s m a x = 9 c m a x / n . - 93 -L e a f z l e v e l _ P h o t o s y n the s i s_Subm The photosynthesis submodel expresses gas concentrations in par t -- 2 - 1 i a l pressure terms, and f lux dens i t ies in umol m s , thereby fol lowing the conventions current ly used by plant phys io log is ts . Chloroplast C0 2 concen-t ra t ions can then be conveniently expressed in p a r t i a l pressure terms (making i t possible to ignore temperature e f fects on gas dens i t y ) , while enzyme a c t i v -i t i e s are treated in the same units as gas d i f f u s i o n conductances. The inputs to t h i s submodel are g e , g . , Q, W, T and barometric pressure, P a . 5 H a In developing the canopy photosynthesis model, i t was found neces-sary to introduce empirical enhancements to those re lat ionships determined by Farquhar and coworkers in the laboratory. Hence the notation used below resembles t h e i r s , but to include the ef fects of var iables not s p e c i f i c a l l y considered by these workers, i t was necessary to introduce some new terms. The maximum potent ia l rate of carboxylation l imi ted by the rate of electron t ransport , in the absence of photorespi rat ion, V . m , „ , i s determined jmax from (Caemmerer and Farquhar, 1981; Farquhar and Caemmerer, 1982; Kirschbaum and Farquhar, 1984): Vjmax " W^W + 2 . U m a x ) 4 . 5 ] (8) where J m a x i s the maximum potent ia l ( l i gh t saturated) rate of whole chain - 2 - 1 electron t ransport , expressed in umol [e lectron] m s , and the factor 4.5 converts i t to the equivalent net mole rate of carboxylat ion. This equation imposes a l i g h t dependent l i m i t a t i o n to photosynthesis, which acts by r e s t r i c t i n g the regeneration of r ibulose diphosphate (RuP 2) substrate and hence the rate of carboxylat ion. - 94 -Analys is of the measured data indicated that afternoon photosyn-thes is rates were lower than morning rates , for condit ions that were otherwise s i m i l a r . This suggested that time since dawn should be considered a factor also inf luencing the photosynthetic processes, and a simple empirical cor rect -ion s i m i l a r to the one used for stomatal conductance was introduced. S i m i l a r -l y , decreasing s o i l water storage was considered to reduce the electron t rans -port l imi ted potent ia l rate of carboxylat ion, V . . Although the evidence for a s i g n i f i c a n t d i rec t e f fect of decreasing W on electron transport (and carb-oxylase a c t i v i t y ) in coni fer fo l iage i s uncertain and possibly contradictory (Beadle and Jarv is 1977), there i s much evidence to suggest that low water potent ia ls do reduce these parameters (Boyer 1976; Pearcy 1983). The function g (W) shown in F i g . 2.1b only produces a s i g n i f i c a n t reduction in g at the »* c lowest values of W recorded at the s i t e , and consequently an equation of th is form was also considered sui table for estimating the ef fects of W on V . : J V j = V j m a x ( 1 " 0 .045t ) (0 .45W 0 * 1 7 ) (9) where t i s the time in hours since dawn, and W i s the water storage in mm. S i m i l a r l y , the apparent maximum rate of carboxy lat ion, V c , with non - l imi t ing RuP2 and C0 2 ( i . e . carboxylase enzyme l imited) i s estimated from (Kirschbaum and Farquhar 1984): V c = (1 - 0 . 0 4 5 t ) ( 0 . 4 5 W ° * 1 7 ) V c m a x / ( l + k . ^ . ) (10) where v c m a x i s the maximum (RuP 2 -saturated) rate of carboxylation for f u l l ac t i va t ion of RuP 2 -carboxylase-oxygenase, (Rubisco), and p c - i s the i n t e r c e l -l u l a r C0 2 p a r t i a l pressure. The Michaelis-Menten constant for ac t i va t ion of Rubisco, k , under given condit ions of pH and M g 2 + concentrat ion, i s here - 95 -assumed constant fo l lowing Kirschbaum and Farquhar (1984), although th is i s considered a s i m p l i f i c a t i o n (Caemmerer and Farquhar, 1984, c i ted in Kirschbaum and Farquhar, 1984). Kirschbaum and Farquhar (1984) found that both RuP 2 - regeneration capacity and the carboxylation rate in E u c a l y p t u s p a u c i f l o r a were temper-ature dependent. It was found that including s i m i l a r temperature functions in the canopy photosynthesis model s i g n i f i c a n t l y improved agreement between measured data (presented in Chapter I) and modelled values. The functions used were of the form suggested by Jones (1983), i . e . : 2 b l ( T + b 2 ) 2 ( T m a x + b2) - (T + b , ) 4 <Tmax + b2> where and b 2 are constants, T m a x i s the temperature at which the maximum reaction rate occurs, and x i s the m u l t i p l i c a t i v e factor appl ied to V. or V . j c Tm a v w a s taken to be 22.7 °C for V. and 21.3 °C for V „ . The values used for max j C b, and b0 respect ively were 1.31 and 4.0 for V. and 1.0 and 5.5 for V . Temperature s e n s i t i v i t y of r*, which i s the C0 2 compensation point in the absence of photorespiratory C0 2 evolution in the l i g h t , was estimated using the equation provided by Brooks and Farquhar (1985) for spinach (Spinaola oleracea): = 4.27 + 0.168(T - T p e f ) + 0.0012(T - T p e f ) 2 (12) where r * i s expressed in Pa and T p g f i s a reference temperature found by t r i a l and error to be about 28 °C for maximum agreement with the measured data (compared to 25 °C suggested by Brooks and Farquhar (1985) for spinach). Following Kirschbaum and Farquhar (1984), the electron transport l imi ted rate of carboxylat ion, W., i s then determined using: - 96 -Wj = V ./ ( l + (7/3)Vpc1) (13) where p . . was i n i t i a l l y assumed equal to the ambient p a r t i a l pressure, p , (see below), while the RuP 2 -saturated rate of carboxylat ion, W c , i s given by: W c = V c / ( l + k c / p c . ) (14) where k £ i s the e f fec t i ve Michaelis-Menten constant for C0 2 ( i . e . regarded as substrate for Rubisco) a f te r taking competitive i n h i b i t i o n by oxygen into account. The gross photosynthesis ra te , P , i s then determined from the minimum of the two calculated carboxylation rates , Wc and W^, using: P g = m1n1mum[Wc,Wj} (1 - V P c 1 ) (15) Stand resp i rat ion i s broken down into several terms, handled at d i f f e r e n t l e v e l s . The photosynthetic demand function equations account for photorespirat ion (1%, being treated as a function of temperature as above), and hence the P g ca lculated by (15) has had photorespirat ion deducted. Growth resp i ra t ion (R g) i s approximated as a f ixed f rac t ion of net photosynthesis, P . Jones (1983) c i tes McCree (1970), who indicated that over a 24 h period Rg can be 0.25-0.34 P , (implying i t may be as much as 0.5 P n ) . Jones also presents values for dark resp i rat ion ( R d ) , in i l luminated photosynthetic c e l l s , in the range 0.05-0.15 P n > Since R d i s the sum of R g and a maintenance resp i rat ion term, Rm , the l a t t e r data imply R g may be less than 0.05 P n . Ludlow and Ja rv i s (1971) reported rates for R d in the range 0.035-0.09 P n for S i tka spruce (Plcea sltchensls). Since these l a s t data are for a c o n i f e r , Rg was estimated as 0.1 P , but i t i s recognised that th i s i s only a rough approximation, and that the se lect ion of a more accurate value (or functional re lat ionship) for R g as a proportion of P n remains to be resolved. - 97 -Maintenance resp i rat ion for each canopy layer , R m f , was computed from: Rmf " O V W 1 ^ 1 - ex P [R k (T -T R 0 ) ] } (16) where T i s the ambient temperature and T R Q i s the temperature at which canopy resp i ra t ion appears to become zero, found by inspection of the data to be about 10.0 °C. R~, R m a v and R. are empi r i ca l l y determined constants repre-T max K senting respect ive ly : the proportion of canopy resp i rat ion contributed by the fo l iage (taken to be 50%); the assumed maximum possible canopy respi rat ion ra te ; and a maintenance resp i rat ion c o e f f i c i e n t set to - 0 . 0 4 7 . The factor R f i s needed to d is t ingu ish between photosynthetic and non-photosynthetic ( i . e . non - fo l ia r ) sources of respiratory CO2 because the l a t t e r should not i n f l u -ence p c - . Equation (16) was derived empi r i ca l l y as a means of resolving the di f ference between measured and modelled photosynthesis rates . The negative R k produces a Q 1 Q value that decreases with increasing temperature, which suggests that a l l the components of canopy resp i rat ion cannot be modelled r e a l i s t i c a l l y with a simple maintenance resp i rat ion funct ion . Ludlow and Jarv is (1971) reported Q 1 Q values for S i tka spruce resp i rat ion that increased from about 1.4 to 3.3 as temperature increased from 5 to 30 °C. By comparison (16) resu l ts in Q 1 Q - 2.42 at 20 °C, decreasing to 1.37 at 30 °C. After estimating stand dry biomass as 40 tonnes h a " 1 , R m f at 25 °C approximates a - 1 -1 resp i rat ion c o e f f i c i e n t of 0.0052 kgL"^] kg [dry biomass] d , which though s l i g h t l y lower than values reviewed by Tanner (1981) for a g r i c u l t u r a l crops, i s consistent with the higher proportion of non-photosynthetic respi r ing t issue occurring in a fo res t . Having obtained values for R m - and P . values mf g for R , R. and P were calculated by solv ing three simultaneous equations: - 98 -Rg = 0 .1P n (17a) R d = R m f + R g (17b) P n = Pg - Rrf (170 to obta in : Rd " «mf + °- l pn <19> Equations (8) through (15) are concerned with ca lcu la t ion of the C0 2 "demand funct ion" i . e . the f lux density of C0 2 uptake required to saturate the f o l i a r photosynthetic processes, as l imi ted by Rubisco and RuP2 a v a i l -a b i l i t y , under the environmental condit ions occurring at that t ime. The C0 2 "supply funct ion" i s the net f lux density of C0 2 from the atmosphere to the inside of the leaves, F 1 , calculated from: F l " MPca - p c i > / P a ( 2 0 ) where the to ta l conductance of the canopy layer to C0 2 d i f f u s i o n , g t , i s given by: -2 -1 where g t has units of umol m s (Jones 1983), D c , D y are the binary d i f -fusion c o e f f i c i e n t s in a i r for C0 2 and water vapour respect ive ly , and R i s the universal gas constant. In (21) g g represents the canopy layer mean stomatal conductance to water vapour. Following Wesely et a l . (1978) and K e l l i h e r et a l . (1986), the corresponding canopy layer mean aerodynamic conductance to water vapour, g A 1 , was assumed to or ig inate mainly from within the canopy ( i . e . as leaf and branch boundary layer terms), such that g A 1 * ( L A r A ) - 1 , - 99 -where r A i s the aerodynamic res is tance, determined from wind ve loc i ty measurements as described in Appendix I I I . In th i s way, equal portions of the tota l canopy aerodynamic resistance were a l located to each canopy layer . The major purpose of the photosynthesis submodel i s to f ind a so lut ion s a t i s f y i n g both demand and supply funct ions . This i s achieved i t e r a -t i v e l y , incrementing or decrementing p . for each i t e r a t i o n , using the current d i f ference between P n and F1 to determine sign and magnitude for the next change in p c - . Convergence i s assumed when the d i f ference (P n - F 1 ) i s less than 5.0 ug[C0 2 ] n f 2 s " 1 ( « 0.1 umol[C0 2] m" 2 s " 1 ) . At t h i s po int , the las t calculated values of F1 and R d are converted to mass f lux d e n s i t i e s , and the value of the layer n o n - f o l i a r maintenance resp i rat ion (R m $ ) determined from: Rms " ^ - R f ) R m a x / R f <22> to be returned as outputs from the submodel. I I _ 5 Canop .y__Leve l_Ca lcu la t ip .ns For each hal f hour, g A 1 , g g and F1 are estimated for each 1 m2 of canopy leaf area, using the value of 0 determined from (5) . Canopy conduct-ance, g c , and canopy net photosynthes the values for each layer i such that:  , is ra te , F 1 , are computed as the sum of n 9 C = I 9 s i (23a) c . = 1 51 V v i ( F l i " R m s i ) ( 2 3 b ) 2 where n i s the number of 1 m layer increments in the canopy. Af ter c a l c u l -at ing F 1 , the estimated s o i l respiratory e f f l u x , R , was subtracted (see - 100 -Chapter I) to give the modelled net C0 2 f lux density at the top of the canopy, F c , which i s d i r e c t l y comparable to the measured data. Having obtained a value for the canopy conductance, i t was possib le to estimate canopy evaporation rate , E, using the Penman-Monteith equation as fo l lows : s(R - G - M) + g.Dpc E = D * & (24) L(s + y ( i + g A /g c ) ) where s i s the slope of the saturat ion vapour pressure curve at T, g A i s the canopy aerodynamic conductance, (= r A _ 1 ) , L i s the latent heat of vapor is -at ion of water at T, y i s the psychrometric constant, and R n , G and M are , respect i ve ly , measured values of net i r rad iance , surface s o i l heat f lux dens-i t y and the rate of canopy heat storage expressed on a ground area bas i s . F i n a l l y canopy water use e f f i c i e n c y was estimated from the ra t io of the modelled F c /E , and expressed on a ha l f -hour l y mass percentage basis i . e . as 100(mg[C02] m" 2 s " 1 )/(mg[H20] m" 2 s - 1 ) . - 101 -III__RIsyLis_A___rjiscy_siON _ _ i _ _ F i g . 2.2 shows measured and modelled canopy conductances for the old stand at Dunsmuir Creek, p lotted on a ha l f -hour l y basis for the s i x days reported in Chapter I. In addit ion s i m i l a r data are presented for 4 August 1984, (remarkable for the dramatic change in i r radiance occurring around mid-day), because t h i s day was considered a rigorous test of the model's response to changing i r rad iance . In general the agreement between modelled and measured g c i s good. For the seven days presented, the overal l regression equation i s : W a s = ° ' 7 5 5 9cmod + 1 A 8 1 <25> 2 with an r of 0.755 calculated from 316 ha l f -hour l y data pairs ( i . e . 20 points excluded) where the units of g are mm s " 1 . The model often under-estimates ear ly morning values soon a f te r dawn. It i s possible that these ear ly morning measured conductances are incor rec t , since the energy balance measurement system i s p a r t i c u l a r l y susceptible to error when ava i lab le energy and atmospheric d i f f u s i v i t y are smal l . There i s a lso the p o s s i b i l i t y that some night - t ime dew f a l l was evaporating from the canopy fo l iage causing exag-gerated latent heat f l u x d e n s i t i e s , even though there was no evidence of pre-c i p i t a t i o n from the weather s tat ion data. Of a l l the days studied, 6 and 20 August 1984 are the two where agreement between modelled and measured data i s poorest. So i l water storage was low on both days (72 and 51 mm respect i ve l y ) . On 6 August (F ig . 2 . 2 f ) , the measured conductances are general ly 50% higher than those predicted by the model, although the model i s evidently responding to f luc tuat ions in solar - 102 -25-i • Measured — Model A 20 August 1983 20-Model B -15- -1 CO (mm 10- • • o D) / • # v / • 5 - / ^ ***** - - ^ // • • ^ 1 ••..v\ — *\ \ \ n U . . ••. — - J 1 1 1 1 1 •• 1 1 1 1 1 1 1 1 1 1 1 I 0 4 8 12 16 20 24 H O U R S ( P S T ) Figure 2 .2a . Comparison of measured and modelled canopy conductance for the old stand at Dunsmuir Creek on 20 August 1983 when s o i l water storage (W) was about 128 mm. The s o l i d l i n e given by Model A i s derived as the minimum of four conductance funct ions , whereas the dashed l i n e given by Model B, estimates g as the product of these same four funct ions . - 103 -25 i i i i i i i - 5 H — ' 1— 1 1 — 1 1— 1 —I— 1 —I— 1— I 0 4 8 12 16 20 24 HOURS (PST) Figure 2.2b. Comparison of measured and modelled canopy conductance for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for F i g . 2 . 2 a ] . - 1 0 4 -25 - i 1 1 1 1 — — i 1 i i I I i 0 4 8 12 16 20 24 HOURS (PST) re 2 .2c . Comparison of measured and modelled canopy conductance for the old stand on 24 Ju ly 1984 when W was about 84 mm. [Notation as for F i g . 2 . 2 a ] . - 105 -- 5 - | ' 1 ' 1 ' 1 • 1 ' 1 ' 1 0 4 8 12 16 20 24 HOURS (PST) Figure 2 .2d. Comparison of measured and modelled canopy conductance for the old stand on 29 Ju ly 1984 when W was about 81 mm. [Notation as for F i g . 2 . 2 a ] . - 106 -25 20 i i i • Measured — Model A - - Model B I I I I I I 4 August 1984 co £ E o 15--10 0 -5 4 8 12 16 20 24 HOURS (PST) Figure 2 .2e . Comparison of measured and modelled canopy conductance for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2 . 2 a ] . - 107 -0 4 8 12 16 20 24 HOURS (PST) Figure 2 . 2 f . Comparison of measured and modelled canopy conductance for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2 .2a ] . - 108 -0 4 8 12 16 20 24 HOURS (PST) Figure 2 .2g . Comparison of measured and modelled canopy conductance for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2 .2a ] . - 109 -i r radiance due to the cloudy condit ions. This day was characterised by low D, so i t appears the g c (D) model function i s not responding adequately to these condi t ions . Conversely, D was general ly higher on 20 August 1984 (F ig . 2.2g) , so the model s i g n i f i c a n t l y overestimated the measured conductance. However, i t i s probable that low s o i l water potent ia l was a l i m i t a t i o n to the measured conductance on th is day, so perhaps the g(W) function i s not responsive enough to low W. Obviously, achieving complete agreement of the model with the measured data i s extremely d i f f i c u l t , and l i m i t a t i o n s in the current model are considered to be ind icat ions to where future improvements should be made. Estimation of canopy conductance as a function of Q was most succ-2 essful when using (5) to ca lcu late mean Q for each 1 m layer of f o l i a g e , but i t was found that canopy conductance could be estimated almost as e f f e c t i v e l y from S b and S d at the top of the canopy, or even d i r e c t l y from the measured value of S. [Note the canopy layer PAR submodel was developed pr imar i l y to ca lcu late values of Q for input to the photosynthesis submodel. These values were also passed to the canopy conductance submodel to maintain consistency. ] Ill j.2 Es t i m a t i o n _ o f _ C a n o r j y _ E C o n d u c t a n c e F i g . 2.3 shows the measured canopy evaporation rate (E) compared with evaporation rates estimated from the Penman-Monteith equation using the ha l f -hour l y modelled canopy conductances for the seven days (F ig . 2 .2 ) . As would be expected, agreement between measured and modelled evaporation i s very general ly s i m i l a r to that already observed for canopy conductance. However, agreement on the cloudy days (26 August 1983, and 4 and 6 August 1984) seems - 110 -Figure 2 .3a . Comparison of measured and modelled canopy evaporation (E) for the old stand at Dunsmuir Creek on 20 August when s o i l water storage (W) was about 128 mm. - I l l --25 H ' 1 ' 1 ' 1 ' 1 ' 1 0 4 8 12 16 20 24 HOURS (PST) Figure 2.3b. Comparison of measured and modelled evaporation (E) for the old stand on 26 August 1983 when W was about 119 mm. - 112 -Figure 2 .3c . Comparison of measured and modelled evaporation (E) for the old stand on 24 July 1984 when W was about 84 mm. - 113 -Figure 2 .3d . Comparison of measured and modelled evaporation (E) for the old stand on 29 Ju ly 1984 when W was about 81 mm. - 114 -Figure 2 .3e . Comparison of measured and modelled evaporation (E) for the old stand on 4 August 1984 when W was about 71 mm. - 115 -re 2 . 3 f . Comparison of measured and modelled evaporation (E) for the old stand on 6 August 1984 when W was about 72 mm. - 116 -F i g u r e 2 . 3 g . C o m p a r i s o n o f measured and m o d e l l e d e v a p o r a t i o n (E) f o r t h e o l d s t a n d on 20 A u g u s t 1984 when W was a b o u t 51 mm. - 117 -bet te r , whereas agreement for 24 July 1984 i s noticeably worse, where midday evaporation i s cons is tent ly underestimated by 10-20%. Part of the explanation for these di f ferences i s seen in F i g . 2.4 which shows that when aerodynamic resistance i s low, and evaporation rates are mainly dependent on the convective term of the Penman-Monteith equation, the s e n s i t i v i t y of E on g c (dE/dg c) i s a lso largely dependent on D. When T (and hence D) i s low, the canopy conductance tends to be high ( t y p i c a l l y in the range 10-15 mm s - 1 ) and consequently dE/dg c i s smal l . Therefore a small error in g c resu l ts in only a small error in E, which helps to explain the remarkably close agreement between measured and modelled E seen on 6 August, when cloudy condit ions caused s i g n i f i c a n t f luctuat ions in ava i lab le energy (F ig . 2 . 3 f ) . Conversely, when T and D are high, the s e n s i t i v i t y dE/dg c i s much greater , p a r t i a l l y because D i s higher, and p a r t i a l l y because g £ has also decreased in response to the higher D. The overal l resu l t i s that a small error in g c w i l l create a r e l a t i v e l y large error in calculated E. F i g . 2.4 also shows that for the young stand, where r A would t y p i c a l l y be much higher, the s e n s i t i v i t y dE/dg£ i s both much greater at low stomatal conductances and smaller at high stomatal conductances than for the old stand under s i m i l a r condi t ions . Since T and D normally increase during the day reaching maxima around mid-afternoon, the maximum errors in E w i l l a lso tend to occur at th is t ime. This explains the noticeable divergence between measured and modelled E during the afternoon of 20 August 1984 (F ig . 2.3g) when the model generally severely overestimated the measured conductance (F ig . 2 .2g) , and on 24 July 1984 when the model underestimated g for the main part of the day. - 118 -Figure 2 . 4 . Canopy evaporation rate as a function of canopy conductance, showing the s e n s i t i v i t y to net I r radiance, vapour pressure d e f i c i t and aerodynamic res is tance . Conditions for a l l curves are : R n : 400 W nf 2 and r A : 10 s m - 1 , except where Indicated, (a) D: 0.5 kPa; (b) D: 1.0 kPa; (c) D: 2.0 kPa; (d) same as (b) except R n : 300 W m " 1 ; (e) same as (b) except r A : 50 s n f 1 . Curves (a) to (d) represent t yp ica l day-time values for the old stand while (e) shows some typ ica l day-time values for the young stand. - 119 -E _ _ l _ _ t i o n _ o f _ t h e _ C a n _ _ _ _ N e F i g . 2.5 shows measured and modelled canopy net photosynthesis for the same seven days shown in F igs . 2.2 and 2 . 3 . The overal l regression equation was: F ^ m o a c • 0 . 9 8 F . m . + 48.2 (26) cmeas cmod 1 ' 2 with an r of 0.515 calculated from 262 data pairs ( i . e . 98 points excluded) - 2 - 1 where the units of F £ are ug m s Most of the excluded points were night - t ime measurements which were disregarded because they were c l e a r l y unreasonable. (The measurement problems caused by the IRGA overranging due to low night - t ime d i f f u s i v i t i e s have already been discussed in Chapter I ) . In order to achieve the level of agreement shown by (26), i t was found necessary to reduce the magnitudes of J „ and V m , from the i r max cmax - 2 - 1 - ? - 1 respective values of 188 umol [e lectron] m s and about 100 umol m s reported by Caemmerer and Farquhar (1981) and Kirschbaum and Farquhar (1984). D iv is ion by 2 had r e l a t i v e l y l i t t l e e f fect on the degree of overestimation by the model, but d i v i s i o n by 4 had a very marked e f f e c t . However, decreasing 2 these values a lso resulted in s l i g h t reductions in the r for measured versus modelled data . Values approximately one t h i r d those reported above were therefore selected as being the best compromise between good agreement in magnitude and good cor re la t ion of measured and modelled data . F i g . 2.5 shows modelled data assuming values of J „ and V „ m „ w reduced to one t h i r d and one max cmax f i f t h of those in the l i t e r a t u r e . - 120 -I t was a lso observed that the model was considerably more sens i -t i ve to changes in J m a „ than in V „ m a v , implying that l i g h t was the major max cma x l i m i t i n g f a c t o r . However, proportionate reductions in J m a v and V „ , v were ma x cma x considered appropriate because studies by Jones (1973, 1985) suggest that when subjected to accumulating water s t r e s s , the " r e l a t i v e l i m i t a t i o n s " of most photosynthetic processes, including the l i g h t react ions , carboxylase a c t i v i t y and stomatal a c t i v i t y , remain v i r t u a l l y constant. The need for reductions in Jm 3 v a n d V „ m = , v could be related to the poor n u t r i t i o n of the stand (Appendix max cma x IV) , since low f o l i a r concentrations of nitrogen (and other nut r ients , p a r t i c -u l a r l y i ron (Bal lard 1987, personal communication)) should be related to the quantity of enzyme s i t e s ava i lab le for carboxylation and other photosynthetic processes. Furthermore, i t appears that Farquhar and coworkers normally used plant material of high nu t r i t i ona l status in the i r work. Appropriate values for J m a „ and V „ „ , „ cannot be determined e a s i -n\a x cma x ly in the f i e l d , a major problem recognised by Ja rv i s e t a l . (1985) and Klippers and Schulze (1985). The canopy photosynthesis model presented here is undoubtedly imperfect, but i t does incorporate equations that r e a l i s t i c a l l y interpret the current state of physiological knowledge. A framework has been constructed, and i f correct values for these biochemical "constants" were determined in the f i e l d or laboratory , they could be incorporated into the model very e a s i l y . The unsat isfactory cor re la t ion between measured and modelled values i s due, at least p a r t i a l l y , to uncertainty in the measured data which would make close agreement with any model extremely d i f f i c u l t . The observed var iat ions in the canopy photosynthesis rates could be real responses to v a r i -ables that have not been considered in the model, so there i s no j u s t i f i c a t i o n for re ject ing the data merely because the agreement i s poor. It i s possible - 121 -C M CO 1000 750 --500--250--u -250 -500 4 -750 i i i i i i i i 20 August 1983 • Measured — Model A •- - Model B 8 12 16 HOURS (PST) 20 24 Figure 2 . 5 a . Comparison of measured and modelled canopy net photosynthesis for the old stand at Dunsmuir Creek on 20 August when s o i l water storage (W) was about 128 mm. Model A used values of V and J . , . . . . . cmax max approximately one t h i r d of those reported in the l i t e r a t u r e by Farquhar and coworkers, whereas Model B used values of approximately one f i f t h . - 122 -Figure 2.5b. Comparison of measured and modelled canopy net photosynthesis for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for F i g . 2 .5a ] . - 123 -1000 -1 1 i r—— i 1 r 1 1 "i 1 r—— • Measured 24 July 1984 — Model A 750-I---Model B CN CO 500 250--8 12 16 20 HOURS (PST) Figure 2 .5c . Comparison of measured and modelled canopy net photosynthesis for the old stand on 24 Ju ly 1984 when W was about 84 mm. [Notation as for F i g . 2 . 5 a ] . - 124 -Figure 2.5d. Comparison of measured and modelled canopy net photosynthesis for the old stand on 29 July 1984 when W was about 81 mm. [Notation as for Fig. 2.5a]. - 125 -1000 • Measured — Model A 750 4 - -Model B 500 4 C M CO o -750 i i i 1 i i I I I 4 August 1984 8 12 16 20 24 HOURS (PST) Figure 2 .5e. Comparison of measured and modelled canopy net photosynthesis for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2 .5a ] . - 126 -1000 - | - i 1 1 i 1 1 1 1 1 1 r • Measured 6 August 1984 — Model A 750- - -Model B -500-- . -750-| ' 1 ' 1 ' 1 ' 1 ' 1 ' 1 0 4 8 12 16 20 24 HOURS (PST) Figure 2 .5 f . Comparison of measured and modelled canopy net photosynthesis for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2 . 5 a ] . - 127 -1000 * Measured — Model A 750 -+-- Model B 500 CN CO 250 u -250---500---750 i i i i i i i i i i 20 August 1984 i — ' — I — ' — I — 1 — h -8 12 16 20 HOURS (PST) 24 Figure 2 .5g. Comparison of measured and modelled canopy net photosynthesis for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2 .5a ] . - 128 -that there i s more error in the data than has been supposed, but th i s would have to be r e s t r i c t e d to the i n f r a - r e d gas analyser (IRGA) measurements, because the energy balance data discussed in Chapter I are general ly very cons istent . Routine checks on the IRGA's performance indicated that C0 2 concentration d i f f e r e n t i a l s were measured r e l i a b l y , while reversal of the a i r intakes every 15 minutes ensured that most systematic errors were el iminated. (Hal f -hour ly measurements of C0 2 concentration d i f f e r e n t i a l known to be spur-ious were corrected or deleted as part of the analys is procedure.) Since the measurements of H and LE were general ly less var iable than the concurrent measurements of A F C , a possible cause of var ia t ion could be short-term f l u c -tuations in average wind d i rec t ion (of the order of 30 minutes). This might be expected because t ransp i rat ion i s e s s e n t i a l l y a physical process (governed by the energy ava i lab le from radiat ion and the l e a f / a i r temperature d i f f e r -ence) whereas photosynthesis i s more phys io log ica l l y determined and hence more sens i t i ve to spat ia l inhomogeneity in leaf area and n u t r i t i o n , and in species composition. The concentration gradient of C0 2 in a i r passing the sensors would therefore probably be more var iable than the comparable gradients in T and e. I t i s evident that the canopy net photosynthesis model responds r e a l i s t i c a l l y to seasonal changes in s o i l water storage and to diurnal meteor-o log ica l v a r i a t i o n s . When W i s high (F ig . 2 .5a) , modelled F c i s higher than on a s i m i l a r day when W i s much lower (F ig . 2 .5g) . When the day i s cloudy, F c responds to changes in i r radiance (F igs . 2.5b and 2 . 5 f ) , whereas on c lear days the diurnal trace of F i s smooth. When temperatures are high (F ig . w 2 . 5 c ) , F c i s general ly lower (and more often negative) than on comparable cooler days (F ig . 2 .5d) . I t i s also quite apparent that the response to changes in temperature and canopy conductance both contribute to the charac-- 129 -t e r i s t i c diurnal pattern where F i s normally highest in the ear ly morning and decl ines throughout the remainder of the day. On 4 and 6 August 1984, when the afternoons were cloudy and c o o l , the model cor rect l y predicts F £ remaining high (F igs . 2.5e and 2 . 5 f ) , although i t tends to underestimate the measured data in mid to la te afternoon (seen p a r t i c u l a r l y in F i g . 2 .5b) . In general Model A overestimates the measured net photosynthesis ra tes , although Model B (with J m a „ and V „ m a „ set to one f i f t h the values max cmax reported in the l i t e r a t u r e ) does t h i s to a lesser extent. 6 August 1984 i s a notable exception when the day-time agreement with Model A i s very good (F ig . 2 . 5 f ) . Consequently the integrated to ta l s for day-time net photosynthesis are also general ly high. However t h i s e f fect i s compensated by a tendency to overestimate the night - t ime canopy resp i rat ion rates , so that on a 24 hour b a s i s , the discrepancy between measured and modelled data i s r e l a t i v e l y small (Table 2 . 1 ) . The model functions are i l l u s t r a t e d in F i g . 2.6 which c lea r l y shows the r e l a t i v e i n s e n s i t i v i t y of F to g except when the l a t t e r becomes small (g £ < 2 mm s " 1 ) . This explains the obvious di f ferences in the diurnal courses of g £ ( F ig . 2.2) and F c ( F ig . 2 . 5 ) . F ig 2.6 a lso shows how the modelled net photosynthesis rate i s c l e a r l y reduced by high temperature e f fec ts on J m , „ and V „ m „ . (Note t h i s i s independent of the modelled ef fects max cmax of temperature on resp i rat ion rate . ) Hence the a p p l i c a b i l i t y of the Farquhar et al. phys io log ica l ly -based approach to modelling forest canopy net photo-synthesis i s demonstrated, in spi te of considerable s i m p l i f i c a t i o n in dealing with l i g h t penetration and obvious shortcomings in the qua l i t y of agreement between predicted F c and the measured data presented here. Table 2.1 Summary of 24 hour total canopy evaporation, net photosynthesis and water use e f f i c i e n c y , measured and modelled, for the old stand on seven days at the Dunsmuir Creek s i t e . DATE Evaporation Net Photosynthesis Water Use YY/MM/DD kg nf g nf mass percentage Measured Modelled Measured Modelled Measured Modelled 83/08/20 2.76 3.60 -6 .04 5.65 -0 .219 0.157 83/08/26 2.22 2.87 3.58 2.43 0.162 0.085 84/07/24 4.38 4.53 -14.59 -24.60 -0 .333 -0 .543 84/07/29 3.68 4.51 -1 .73 - 3 . 1 6 -0 .047 -0 .070 84/08/04 2.54 3.22 0.44 0.30 0.017 0.009 84/08/06 2.71 2.73 3.13 3.85 0.115 0.141 84/08/20 2.23 3.62 -8 .23 1.89 -0 .369 0.052 - 131 -Figure 2 .6a . S e n s i t i v i t y of canopy net photosynthesis model to changes in so lar i r rad iance , and stomatal conductance, at 12:15 PST on 24 July 1984, a hot, c lear day. The s o l i d curves represent the C0 2 demand function for each canopy layer , where the uppermost l i n e represents the top layer . Lower l i nes indicate the e f fec ts of reduced electron t rans -port rates in lower canopy layers due to l i g h t absorption by upper layers . The dashed l ines represent the C0 2 supply functions corre-sponding to measured ( r ight) and modelled ( l e f t ) mean layer g g . The in te rsect ion points enable the layer C0 2 f l ux dens i t ies to be estimated (excluding "day-time dark r e s p i r a t i o n " , R d ) . Conditions were as fo l lows : S: 876 Wm , T: 30.0 °C, measured g : 4.36 mm s " 1 , modelled g c : 3.36 mm s " 1 , modelled R d : 107 ug m" 2 s per canopy layer , e s t i -mated s o i l respi ratory C0 2 e f f l u x , R $ : 100 ug m s - 1 . - 132 -Figure 2.6b. S e n s i t i v i t y of canopy net photosynthesis model to changes in so lar i r rad iance , and stomatal conductance, at 12:15 PST on 6 August 1984, a predominantly cool and cloudy day. Conditions were as fo l lows: S: 709 W m " 2 , T: 17.3 °C, measured g „ : 13.6 m m s - 1 , modelled g : 9.8 mm s , modelled R^: 57 p g n s per canopy layer , estimated s o i l respi ratory C0 2 e f f l u x , R $ : 100 ug n f 2 s " 1 . [For further explan-at ion see a lso F i g . 2 .6a ] . - 133 -121^4 Es t i m a t i o n _ o f _ C a n o r j y _ W a t M o d e l l e d _ C a n o £ y _ E v a £ o r a t i o n F i g . 2.7 shows measured and modelled day-time water use e f f i c iency ra t ios (WUE) for the seven days. (Night-time values have been omitted because the measured values are subject to considerable scatter and are general ly two to three orders of magnitude larger than in day-t ime, as seen in Chapter I, F igs . 1 .8 -1 .10 . ) I t can be seen that when day-time WUE are considered in i s o l a t i o n , the agreement between measured and modelled values i s ac tua l l y better than for photosynthesis and evaporation separately . This i s because of the tendency of the submodels to overestimate both E and F c , although WUE i s s t i l l generally s l i g h t l y overestimated by the model. The obvious exceptions to th i s are the cloudy days (F igs . 1.7b, 1.7e and 1 . 7 f ) , when mid to la te afternoon measured WUE usual ly s i g n i f i c a n t l y exceeds the modelled values, because F i s under-estimated at these times (see above). Farquhar and Sharkey (1982) show that optimal water use w i l l occur when photosynthesis rate occurs at the intersect ion of the carboxy lase -act i -v i t y l imi ted and the l i g h t - l i m i t e d demand funct ions , i . e . when the supply function crosses the break-point in the demand curve. F i g . 2.6 suggests that for the old stand canopy, water use was greater than opt imal , at least at mid-day for the two days presented, since the supply function intersects with the l i g h t - l i m i t e d port ion of the demand curve. Under these condit ions a given increase in conductance would resul t in a r e l a t i v e l y large increase in t rans -p i ra t ion but only a small increase in net photosynthesis. The reasons for t h i s might be related to the poor canopy n u t r i t i o n , since there i s some good evidence that for c o n i f e r s , t h i s can lead to poor stomatal control and hence reduced water use e f f i c i e n c y (Schomaker 1969; Bengtson and Voigt 1962). - 134 -0 4 8 12 16 20 24 HOURS (PST) Figure 2 .7a . Comparison of measured and modelled canopy day-time water use e f f i c i e n c y r a t i o for the old stand at Dunsmuir Creek on 20 August 1983 when s o i l water storage (W) was about 128 mm. Night-t ime values were omitted because low atmospheric d i f f u s i v i t i e s and low evaporation rates combined with large respiratory f luxes resulted in very large and e r r a t i c r a t i o s . - 135 -0 4 8 12 16 20 24 HOURS (PST) Figure 2.7b. Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 26 August 1983 when W was about 119 mm. [Notation as for F i g . 2 . 7 a ] . - 136 -HOURS (PST) Figure 2.7c. Comparison of measured and modelled canopy day-time water use efficiency ratio for the old stand on 24 July 1984 when W was about 84 mm. [Notation as for F ig . 2.7a]. - 137 -0 4 8 12 16 20 24 HOURS (PST) Figure 2 .7d . Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 29 Ju ly 1984 when W was about 81 mm. [Notation as for F i g . 2 .7a ] . - 138 -Figure 2.7e. Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 4 August 1984 when W was about 71 mm. [Notation as for F i g . 2 .7a ] . - 139 -HOURS (PST) Figure 2 . 7 f . Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 6 August 1984 when W was about 72 mm. [Notation as for F i g . 2 .7a ] . - 140 -0 4 8 12 16 20 24 HOURS (PST) Figure 2 .7g . Comparison of measured and modelled canopy day-time water use e f f i c i e n c y ra t io for the old stand on 20 August 1984 when W was about 51 mm. [Notation as for F i g . 2 .7a ] . - 141 -Balanced against t h i s proposit ion i s the fact that the measured WUE rat ios were general ly s l i g h t l y smaller than the modelled values, which might indicate that overal l canopy WUE i s ac tua l l y c loser to opt imal . The canopy model demonstrates the p o s s i b i l i t y of predict ing annual forest product iv i ty from basic cl imate data alone, although the one presented here i s imperfect. A l t e r n a t i v e l y , the agreement between measured and modelled WUE over 24 hours (Table 2.1) suggests the model might be used successful ly to obtain annual estimates of net photosynthesis from WUE determined as a func-t ion of W. The evident increase in WUE as W decreased (reported in Chapter I) indicates that WUE changes pred ic tab ly , so i t may be possible to estimate annual photosynthesis from modelled E using approaches analogous to those proposed by Tanner (1981) and reviewed by Tanner and S i n c l a i r (1983) for a g r i c u l t u r a l crops. 11__5 Cr i . t j . _ue In p r i n c i p l e , the phys io log ica l ly -based model of canopy net photo-synthesis was found to work, although less successfu l ly than had been hoped. Af ter se lect ing su i table values of the physio logical parameters J and max V cmax ( C a e m m e r e r a n d Farquhar 1981), about one th i rd the magnitude of those reported in the l i t e r a t u r e of Farquhar and co-workers, i t was possible to explain only s l i g h t l y more than 50% of the var ia t ion in the measured data. A l l d r i v ing var iables l i k e l y to be s i g n i f i c a n t were considered in the model, while in terpretat ion of the i r e f fects used some of the most recently reported physio logical re la t ionsh ips . It i s possible that a more empir ical model would have been more successfu l . For example, the model of Klippers and Schulze (1985) produced extremely good agreement with measurements obtained from a leaf cuvette system. However the i r greater success could also indicate the - 142 -necessity for a better submodel of canopy l i g h t penetration in the model presented here. Ja rv i s et al. (1985) and Landsberg (1981) both suggest that a good c r i t e r i o n for assessing the u t i l i t y of a model i s whether i t allows the real values to be predicted at least as r e l i a b l y as they can be measured. Given the l i m i t a t i o n s of the energy balance/Bowen Ratio technique, i t i s reasonable to expect that LE and F c can be measured to at least ±20% during the day-t ime. Since the model of canopy conductance only explained about 75% of the v a r i a b i l i t y , i t i s barely acceptable. I f conductance could be p red ic t -ed more success fu l l y , then canopy evaporation estimated from the Penman-Monteith equation would also be more accurate. Probably a more sophist icated canopy l i g h t penetration submodel, such as that of Norman and Jarv is (1975), would p a r t i a l l y resolve the discrepancy between measured and modelled canopy conductances. Even Norman's simpler models discussed by Jarv is and Leverenz (1983) or that of Landsberg et al. (1970) might be expected to provide better r e s u l t s . The minimum require -ments would be some determinations of average leaf angle and the degree of clumping of fo l iage around tree branches and stems. Cer ta in ly a better s imul -at ion of canopy l i g h t penetration i s required to better estimate canopy net photosynthesis (where only 50% of the v a r i a b i l i t y was explained by the model), which probably explains why the model of Ja rv i s et al. (1985) resulted in s i g n i f i c a n t l y greater success than in t h i s study, even though they used a s impler , empir ical photosynthetic demand funct ion . A further reduction in the discrepancies between measured and modelled data might be achieved by t reat ing the understory and overstory canopies as separate e n t i t i e s , along the l ines of K e l l i h e r et al. (1986). - 143 -The canopy conductance c h a r a c t e r i s t i c s , the values of J m a u and V _ _ _ „ , and max cma x the canopy ex t inc t ion coe f f i c ien ts could a l l be determined separately for tree and sa la l vegetation components. Suitable values for some of these data might be obtained from the l i t e r a t u r e , or from cooperative studies with plant phys io log ists and biochemists. Another major weakness in the current photosynthesis model prob-ably contr ibut ing to poor agreement with the measured data , i s the inadequate treatment of s o i l and canopy r e s p i r a t i o n . A separate study devoted to measur-ing tree resp i rat ion such as that of Ninomiya and Hozumi (1981), should enable much better re lat ionships between T and R d to be estab l i shed , and also deter -mine the r e l a t i v e contr ibut ions of the maintenance and growth respiratory terms. A better model of s o i l respiratory e f f lux would a lso help reduce the uncertainty in ca lcu la t ing the f i n a l carbon balance. - 144 -IV CONCLUSIONS A simple empirical model of canopy conductance in a young Douglas-f i r stand, based on four environmental var iables was found to be moderately successful in predict ing the diurnal var ia t ion in canopy conductance and 2 evapotranspirat ion. The r of about 0.75 indicated that the model predicted the true conductances almost as well as they could be measured. However, even r e l a t i v e l y large discrepancies between measured and modelled canopy conduct-ance were general ly found to resul t in only s l i g h t errors in day-time canopy evaporation estimated using the Penman-Monteith equation. The largest errors in modelled canopy evaporation occurred when stomatal conductance was reduced - p a r t i c u l a r l y in response to high atmospheric vapour pressure d e f i c i t . A complex phys io log ica l ly -based model of net canopy photosyn-t h e s i s , derived from the recent work of Farquhar and coworkers, was found to be only p a r t i a l l y successful in predict ing the diurnal var ia t ion in measured 2 canopy net C0 2 f lux density when applied to the same stand canopy. The r of about 0.5 suggested that in i t s present form, the model could not predict the true f luxes to the same accuracy with which they were measured using a modi-f i e d energy balance/Bowen Ratio technique. However, the model successful ly predicted trends in diurnal net photosynthesis and for seven days studied in d e t a i l , predicted 24 hour to ta l net photosynthesis f a i r l y c l o s e l y . Errors in canopy conductance noted above were not normally of great s ign i f icance in estimating canopy net photosynthesis, because under most of the studied condit ions the net C0 2 f lux density was evidently demand-limited rather than s u p p l y - l i m i t e d . Solar i r radiance was almost invar iab ly the l i m i t i n g factor for net photosynthesis even when conductance was low and a i r temperature very high. - 145 -Possible reasons for the observed discrepancies were: (1) inade-quate simulation of the d i s t r i b u t i o n of photosynthet ical ly act ive radiat ion wi th in the canopy, (2) lack of d i s t i n c t i o n between the photosynthetic and stomatal c h a r a c t e r i s t i c s of the overstory (trees) and understory (sa la l vege-t a t i o n ) , and (3) measurement error being greater than the ±20% estimated from error a n a l y s i s . Before the model can be considered of value in explaining growth responses to short term var ia t ion in weather, and hence form a component of a higher level stand product iv i ty model, several s i g n i f i c a n t improvements to the photosynthesis model are required. These should concentrate on three areas. F i r s t l y , and most importantly, the means by which f o l i a r l i g h t environment is predicted for any level in the canopy. Secondly, the values used for maximum f o l i a r electron transport ra te , J m a x . and for maximum carboxylase a c t i v i t y , v , - m a „ > should be determined from f i e l d and laboratory measurements, for both overstory and understory components of the canopy. Th i rd l y , better measure-ments and understanding of s o i l and canopy respiratory parameters are required to. enable more accurate modelling of the respiratory contr ibut ions to canopy level CO,, f lux d e n s i t i e s . - 146 -V REFERENCES Bare, B.B. 1986. Douglas - f i r y i e l d forecast ing systems: what's a v a i l a b l e . In C D . O l i v e r , D.P. Hanley and J .A . Johnson (eds . ) , Doug las - f i r : Stand Management for the Future. Proceedings of symposium held at the Univers i ty of Washington, 18-20 June 1985. pp. 344-349. Beadle, C.L . and P.G. J a r v i s . 1977. The ef fects of shoot water status on some photosynthetic p a r t i a l processes in S i tka spruce. P h y s i o l . P lant . 41: 7 -13 . Beadle, C . L . , P.G. J a r v i s , H. Talbot and R.E. Nei lson . 1985. Stomatal conductance and photosynthesis in a mature Scots pine fo res t : I I . Dependence on environmental var iables of s ingle shoots. J . Appl . Eco l . 22: 573-586. Bengtson, G.W. and G.K. Voigt . 1962. A greenhouse study of re lat ions between nutr ient movement and conversion in a sandy s o i l and the n u t r i t i o n of slash pine seedl ings. So i l S c i . Soc. Amer. Proc. 26 :609-612 Blackman, F.F. 1905. Optima and l i m i t i n g fac to rs . Ann. Bot. 19: 281-295. Boyer, J . S . 1976. Photosynthesis at low water p o t e n t i a l s . P h i l . Trans. R. Soc. Lond. B 273: 501-512. Br istow, K.L. and G.S. Campbell. 1985. An equation for separating da i l y so lar i r r a d i a t i o n into d i rec t and d i f fuse components. Ag r i c . For. Meteor. 35: 123-131. Brooks, A. and G.D. Farquhar. 1985. Ef fect of temperature on the C 0 ? / 0 ? s p e c i f i c i t y of r ibu lose - l ,5 -b isphosphate carboxylase/oxygenase and the rate of resp i rat ion in the l i g h t . Estimates from gas-exchange measurements on spinach. Planta 165: 397-406. Caemmerer, S . von, and G.D. Farquhar. 1981. Some re lat ionships between the biochemistry of photosynthesis and the gas exchange of leaves. Planta 153: 376-387. Campbell, G.S. 1985. So i l Physics with BASIC. Transport Models for S o i l -Plant Systems. E l s e v i e r , Amsterdam. 150 pages. Campbell, G.S. 1977. An Introduction to Environmental Biophysics. Springer-Ver lag, New York. 159 pages. Charles-Edwards, D.A. , D. Doley and G.M. Rimmington. 1986. Modelling Plant Growth and Development. Academic Press, Sydney, A u s t r a l i a . 233 pages. Cowan, I.R. 1982. Regulation of water use in re la t ion to carbon gain in higher p lants . In O.L. Lange, P.S. Nobel, C.B. Osmond and H. Z iegler (eds . ) . Phys io logical Plant Ecology I I . Water Relat ions and Carbon A s s i m i l a t i o n . Encyclopedia of Plant Physiology, new Ser. V o l . 12B. Spr inger -Ver lag , B e r l i n , pp 589-613. - 147 -C u r t i s , R.O. , G.W. Clendenen and D.J . DeMars. 1981. A new stand simulator for coast Doug las - f i r : DFSIM user 's guide. USDA For. Serv. Gen. Tech. Rep. PNW-128. Pac. Northwest For. and Range Exp. S t n . , Por t land, OR. 79 pages. Farquhar, G.D. and S. von Caemmerer. 1982. Modelling of photosynthetic response to environmental condi t ions . In O.L. Lange, P .S. Nobel, C.B. Osmond and H. Z ieg ler (eds . ) . Physiological Plant Ecology I I . Water Relations and Carbon A s s i m i l a t i o n . Encyclopedia of Plant Physiology, new Ser. V o l . 12B. Spr inger -Ver lag, B e r l i n , pp 549-587. Farquhar, G.D. and T.D. Sharkey. 1982. Stomatal conductance and photosyn-t h e s i s . Ann. Rev. Plant P h y s i o l . 33: 317-345. Farquhar, G.D. and S.C. Wong. 1984. An empirical model of stomatal conduct-ance. Aust. J . Plant P h y s i o l . 11: 191-210. Gates, D.M. 1965. Radiant energy, i t s receipt and d i s p o s a l . Meteor. Monogr. 6: 1-26. Gates, D.M. 1980. Biophysical Ecology. Spr inger -Ver lag , New York. 611 pages. Go l lan , T . , J . B . Passioura and R. Munns. 1986. Conductance of f u l l y turgid wheat and sunflower leaves. Aust. J . Plant P h y s i o l . 13: 459-464. Grace, J . G . , D.C. Malcolm and I.K. Bradbury. 1975. The e f fect of wind and humidity on leaf d i f f u s i v e resistance in S i tka spruce seedl ings. J . Appl . E c o l . 12: 931-940. Hamilton, G . J . and J . M . C h r i s t i e . 1971. Forest Management Tables (Metr ic ) . Forestry C