a05.gif (1361 bytes)   ¥x¤jºØ¤l¬ã¨s«Ç

±MÃD1¡G«í·Å»PºØ¤lµoªÞ³t²v

   2007-03-29    ¦^¥Í²z¾Ç¥Ø¿ý

ºô¸ôºØ¤l¾Ç¬ÛÃöºô­¶

·Å«×»PºØ¤lµoªÞ (·Å«×»PµoªÞ¦Ê¤À²v¡G«í·Å¤U¡F¿n¼ö»PµoªÞ³t²v)

¤ô»PºØ¤lµoªÞ (¤ô¶Õ»PµoªÞ³t²v¡GÄ­¤ô»PµoªÞ³t²v¡B¤ô¼ö¿nÄ­­È»PµoªÞ³t²v)

¦^¥Ø¿ý

¡@

¡@

¡@

¤½¦¡±Àºt              ¦^¥Ø¿ý

q1  at sub-optimum temperature

thermal time(number of degree-days above base temperature) to germination at day t.

q1 = (T1-Tb) + (T2-Tb) +...+ (Tt -Tb) = t(T-Tb)

At constant temperautre,

1/t =(T- T b) /q1 = (-T b /q1 ) + T(1/q1 ) = K + (1/q1 ) T

consider germination of seed fraction G, e.g. 10,20,...,90,100¢M

1/t(G) = [T-Tb (G)]/q1 (G)     (Since there is a common base temperature in each G)

= (T-Tb) /q1 (G) .....................................(1)

Now, G has a normal distribution onq1 (G), with a standard deviation (s) of the frequency distribution of thermal times in the seed population,

probit (G) = K` + q1 (G)/s...............................(2)

q1 (G) = [probit(G)- K`]s.................................(3)

From (1)¡®(3), the relationship between rate of germination and sub-optimum temperature can be described for the whole population by a single equation

1/t(G)= (T-Tb) /{[probit(G)- K`]s }.........................(4)

[probit(G) -K`]s = (T-Tb)t(G)

probit(G) = K + [(T-Tb)t(G) ] /s...........................(5)

q2 at supra-optimum temperature

thermal time(number of degree-days above base temperature)to germination at day t.

1/t = (Tc - T) /q2

1/t(G) = (Tc (G) - T) /q2 (G) ...................................(6)

Now,G has a normal distribution on Tc (G), with a standard deviation ƒvƒãƒw

of the frequency distribution of ceiling temperatures in the seed population,

probit(G)= K`` - Tc (G) /s ...............................(7)

Tc (G)= [K`` - probit(G)]s

¦^¥Ø¿ý

¡@

¡@

¡@

from (6)& (7)

1/t(G) = {{[K`` - probit(G)]s - T} /q2

{[K``- probit(G)]s} - T = [1/t(G)]q2

[K``- probit(G)]s = T + q2 / t(G)

probit(G) = Ks - [T + q2 / t(G)] /s.....................(8)

¦^¥Ø¿ý

¡@

¡@

¡@

¸ÕÅ礶²Ð

(I)Covel et al., 1986

¨ú4ºØ¨§ÃþºØ¤l¡A¦b5-40¢J¤§¶¡¨C¹j5¢J°µ¤@·Å«×³B²z¡A½Õ¬dµoªÞ²v²Ö¶i¡A¨Ì¹Ï¦ôºâ¨C10%(G)µoªÞºØ¤l©Ò»Ýªº®É¶¡[t(G)]¡A§@µoªÞ³t²v1/t(G) vs ·Å«×Tªº¹Ï¡C¥Ø´úTo¡ATo¥H¤UªÌ¨C­ÓG¶i¦æ³t²v1/t(G) vs ·Å«×Tªº°jÂk¤ÀªR(¦¡1)¡A±o¨ì9­ÓTb (G), ¤Îq1(G)¡CµM«á9±ø°jÂk¥O¨ã¦³¬Û¦PªºTb¡AµM«á¦A­pºâ°jÂk¤èµ{¦¡¡A¤Ï´_¥H¤£¦PªºTb (¨C¹j0.5¢J°µ¤@¦¸)¶i¦æ¡A¤ñ¸ûµ²ªG¥H±o¨ì³Ì¤p residual varianceªº§@¬°To¥H¤Uªº¦@¦PTb¡C°w¹ï¦@¦PTb©Î9­ÓTb (G)¶i¦æ®t²§¤ñ¸û¡CTo¥H¤WªÌÃþ¦ü¤èªk¶i¦æ­pºâ¡C---------------------

µ²ªG¡G¦@¦PTb¦¨¥ß¡A¨C²ÕºØ¤lTb¬Û¦P¡Aq1(G)¤£¦P¡C¹ïG vs q1¶i¦æ»s¹Ï¡AµM«á¥H¦@¦PTb©Ò±oq1¶i¦æprobit¤ÀªR¡A°£¤F¤Ö¼Æ¥H¥~¡A¤j­P¾A°t¡A¯S§O¬O10-90%ªº³¡¥÷¡C ¼Ò¦¡4¦¨¥ß¡C

¡@

(II)Ellis et al., 1986

¨úchickpea¤­­Ó«~ºØ¡A¦P«e¦A¶i¦æ¸ÕÅç¤Î¤ÀªR¡A¨Ã¥t¤ÀªR«~ºØ¶¡Tbªº®t²§ÅãµÛ©Ê¡Cµ²ªGÃþ¦ü¡A¥B«~ºØ¶¡Tb¬Û¦P¡F¤j©óTo³¡¥÷«ê¦n¬Û¤Ï¡C

®Ú¾Ú¼Ò¦¡5¥i±æÂ²¤Æ¸ÕÅç¡Cprobit(G) = K + [(T-Tb)t(G) ] /s   ¥Ñ¨â­Ó·Å«×¸ÕÅç©Ò±oªºprobit(G)¡BT¡Bt(G)¸g¥Ñ¤Ï´_¤ÀªR¨D±o¦h²ÕK¡BTb¡Bs¥H³y¦¨³Ì¤presidual variance µø¬°¥¿½T­È¡C¤j©óTo³¡¥÷¤]¬O¦P¼Ë¡C

¡@

(III)Ellis et al., 1987

¨úÅú¨§¤­­Ó«~ºØ¡A¦P«e¦A¶i¦æ¸ÕÅç¤Î¤ÀªR¡C¥t¶i¦æ¥H¤U¸ÕÅç¡AºØ¤l¦b10¡B20¡B27¡B30¢Jµ¥¥|·Å«×¤U¶i¦æµoªÞ¸ÕÅç¡C

To¥H¤U¡G¥H-10¦Ü0¢J¡A¨C0.5¢J¤@¦¸°µ¬°Tb¡A¶i¦æ21¦¸¤ÀªR¡AG¹ï(T-Tb)t(G) §ä¥X³Ì¾ATb¡A¦b¨M©w¤­­Ó«~ºØ¦³µL¦@¦PTb¡C

To¥H¤W¡Gq2    ¨C1¢J¤@¦¸¶i¦æTc¹ïT+q2 /t(G)ªº¤ÀªR¡C¨D±o³Ì¨Îq2 ¡BKs¡Bs¡C

µ²ªG¡GTo¥H¤U¡Gcv. Sutton¡G ­ì¤ÀªR¦¨¥ß¡A¥Ñ¨â·Å«×©Ò±oªº¼Ò¦¡»P¦U·Å«×ªº¼Æ¾Ú¬Û¦X¡CÀ³¥Î¨ì¨ä¥¦«~ºØ¥çµM¡C                          ¦^¥Ø¿ý

¡@

¡@

¦^¥Í²z¾Ç¥Ø¿ý

¡@