Field methods for sampling and storing nectar from flowers with low nectar volumes

Transcription

1 Annls of Botny 13: , 29 doi:1.193/o/mn241, ville online t TECHNICAL ARTICLE Field methods for smpling nd storing netr from flowers with low netr volumes D. S. Morrnt 1, R. Shumnn 2 nd S. Petit 1, * 1 Sustinle Environments Reserh Group, Shool of Nturl nd Built Environments, University of South Austrli, Mwson Lkes, South Austrli 595, Austrli nd 2 Levy & Co Environmentl Servies, In Wrk Reserh Institute, University of South Austrli, Mwson Lkes, South Austrli 595, Austrli Reeived: 25 August 28 Returned for revision: 8 Otoer 28 Aepted: 3 Novemer 28 Pulished eletronilly: 11 Deemer 28 Bkground nd Aims Although severl methods of smpling nd storing florl netr re ville, little informtion exists on smpling nd storing netr from flowers with low netr volumes. Methods for smpling nd storing netr from the flowers of speies with low florl netr volumes (,1 ml) were investigted using the flowers of Eulyptus speies. Methods Smpling with miropillry tues, lotting up with filter pper, wshing nd rinsing were ompred to determine msses of sugrs reovered nd differenes in sugr rtios. Storge methods inluded room temperture, refrigertion nd freezing tretments; the ddition of ntimiroil gents enzyl lohol or methnol to some of these tretments ws lso evluted. Netr smples were nlysed using high-performne liquid hromtogrphy nd the msses of surose, gluose nd frutose in eh smple were determined. Key Results Msses of sugrs vried signifintly mong smpling tretments, ut the highest yielding methods, rinsing nd wshing, were not signifintly different. A wshing time of 1 min ws s effetive s one of 2 min. Storge trils showed tht the sugr onentrtion mesurements of netr solutions hnged rpidly, with the est results hieved for refrigertion with no dditive (surose nd frutose were stle for t lest 2 weeks). Sugr rtios, however, remined reltively stle in most tretments nd did not hnge signifintly ross 4 weeks for the methnol plus refrigertor nd freezing tretments, nd 2 weeks for the refrigertion tretment with no dditive. Conlusions Wshing is reommended for netr olletion from flowers with low netr volumes in the field (with the understnding tht one wsh underestimtes the mounts of sugrs present in flower), s is immedite nlysis of sugr mss. In view of the gret vrition in results depending on netr olletion nd storge methods, ution should e exerised in their hoie, nd their ury should e evluted. The use of pulsed mperometri detetion, more speifi thn refrtive index detetion, my improve the ury of netr sugr nlysis. Key words: Eulyptus, flower with smll netr volume, netr olletion, netr smpling, netr storge, sugr rtio. INTRODUCTION Methods to mesure the volume, sugr ontents nd energy vlue of netr re used in the study of mny eologil proesses (Dungn et l., 24), in prtiulr the study of plnt niml intertions (Bolten nd Feinsinger, 1978; Kerns nd Inouye, 1993), nd n permit the lultion of rrying pity for netrivores (Petit nd Pors, 1996). Sugr omposition my lso ffet visitor preferene or ssoition (Hinsworth nd Wolf, 1976; Bolten nd Feinsinger, 1978; Mrtínez del Rio et l., 1992; Shondue nd Mrtínez del Rio, 23). Severl methods re ville to smple netr, nd seleting the est one n e diffiult (Lloyd et l., 22). The effetiveness of hosen tehnique is influened y florl morphology, netr hrteristis nd smpling regime (Bolten nd Feinsinger, 1978; Kerns nd Inouye, 1993; Lloyd et l., 22). While mny ppers hve een pulished on the smpling of florl netr, puity of informtion exists regrding smpling from flowers tht produe low netr volumes (,1 ml), nd lso on the onditions under whih the resulting netr smples should e stored. * For orrespondene. E-mil Sophie.Petit@unis.edu.u Lloyd et l. (22) desried qulittive method for visul ssessment of netr volume tht did not involve the removl of netr from flowers; however, this method did not llow for quntittive nlysis of the sugrs in the netr. Methods desried to smple netr from flowers, whih llow for suh nlysis, inlude: () soking up the netr with filter pper wiks (MKenn nd Thomson, 1988; Kerns nd Inouye, 1993; Coret, 23); () drwing up the netr in miropillry tues, mirosyringes or miropipettes (Collins nd Newlnd, 1986; MKenn nd Thomson, 1988; Kerns nd Inouye, 1993; Lnz et l., 1995; Coret, 23; Tshpk, 24); () gitting (wshing) the flower in stoppered tue of distilled wter (Käpylä, 1978; Grünfeld et l., 1989); (d) rinsing flowers with distilled wter (Núñez, 1977; Mllik, 2); (e) entrifuging individul flowers or infloresenes (Swnson nd Shuel, 1949; Armstrong nd Pton, 199); nd ( f ) removing netr with power-driven spirtors (Armstrong nd Pton, 199). The onentrtion of netr solution is importnt in the energetis of forging (Heinrih, 1975). The sugr onentrtion of netr n e mesured in the field using hnd-held refrtometer (Bolten nd Feinsinger, 1978; Bond nd Brown, # The Author 28. Pulished y Oxford University Press on ehlf of the Annls of Botny Compny. All rights reserved. For Permissions, plese emil: journls.permissions@oxfordjournls.org Downloded from y guest on 4 My 218

2 534 Morrnt et l. Smpling nd storing florl netr t low volumes 1979; Collins nd Newlnd, 1986; Kerns nd Inouye, 1993; Mllik, 2; Dungn, et l. 24), or nlysed in lortory using tehniques suh s hromtogrphy or spetrophotometry. High-performne liquid hromtogrphy (HPLC) is n esy nd urte wy to qulify nd quntify speifi sugrs nd other hemils within netr smple (Kerns nd Inouye, 1993). Where netr volumes re smll (,1 ml), nd mesurements of the volume nd onentrtion of tht netr re not required, netr n e lotted up from flowers with smll pre-dried nd weighed wiks of filter pper (MKenn nd Thomson, 1988; Kerns nd Inouye, 1993; Coret, 23). The hnge in mss of the pper fter redrying n provide n estimte of the mss of sugrs in the netr (Dungn et l., 24). The filter pper my lso e soked in solvent, nd the mss nd omposition of solutes n e nlysed using tehniques inluding hromtogrphy or olorimetri nlysis (MKenn nd Thomson, 1988). The filter pper wik method does not, however, provide mesurements of netr volumes (MKenn nd Thomson, 1988). The volume of netr in flowers is limiting ftor in netr intke y netrivores nd should e reported in netr investigtions (Bolten et l., 1979). Miropillry tues nd refrtometers re ommonly used in the field for mesuring, respetively, the volume nd sugr onentrtion of reltively undnt florl netrs (Collins nd Newlnd, 1986; MKenn nd Thomson, 1988; Coret, 23). Netr is removed from flower using fixed-ore tue, nd the volume is mesured y determining the length of the liquid olumn within the tue. The netr is then pplied to refrtometer to mesure the sugr onentrtion (Coret, 23). Miropipettes nd mirosyringes n e used insted of miropillry tues to hieve similr results (Kerns nd Inouye, 1993). While the miropillry tue nd refrtometer methods re useful for smpling the netr of mny plnt speies, they re less effetive where florl netr is highly visous, or is produed in low (,1 ml) volumes (Käpylä, 1978; MKenn nd Thomson, 1988; Mllik, 2; Mnets nd Petropoulou, 2; Coret, 23; Dungn et l., 24; Birthnell nd Gison, 26). All ville netr nnot e extrted with miropillry tues (Stephnou et l., 2; Mnets nd Petropoulou, 2), nd proing my dmge the netry tissue nd introdue ell ontents into the netr (Willmer, 198). Additionlly, it is diffiult to mesure low volumes of netr with hnd-held refrtometers, exept those with extremely sensitive prisms (MKenn nd Thomson, 1988). Even if reding n e otined using hnd-held refrtometer, non-sugr onstituents suh s mino ids n ontriute to the refrtive index nd skew lultions of netr energy ontent sed on sugr onentrtion (Inouye et l., 198). While the netr smpling studies mentioned previously exmine methods of smpling florl netr from flowers, few studies hve exmined the methods of storing these smples nd the impt of teril nd fungl infetion on netr storge life. Freezing (e.g. Lnz et l., 1995) ould e n effetive storge method, lthough it is not lwys prtil method in field sitution, s ould refrigertion (Herert nd Clder., 1983; Erhrdt nd Bker, 199) nd/or the ddition of ntimiroil gents. This study ws onduted to determine the most effetive tehniques for smpling nd storing netr from flowers with low netr volumes. The informtion otined from these studies will e of use to reserhers investigting the sugr ontents of the netr of plnt speies with low florl netr volumes. Plnt speies used MATERIALS AND METHODS Methods for the olletion nd storge of florl netr were ompred using flowers from four speies of Eulyptus flowering on the grounds of the Mwson Lkes mpus of the University of South Austrli. Beuse few flowering trees were ville for eh of the three experiments (wshing durtion, smpling nd storge), different speies of eulypt ws used for eh. The flowers of yellow gum (Eulyptus leuoxylon ssp. leuoxylon) were used to test the effets of gitting nd soking flowers for different lengths of time. Swmp mllet (Eulyptus spthult) flowers were used for the omprison of smpling methods. The flowers of orl gum (Eulyptus torqut; Fig. 1) were used to test storge of netr solutions, whih were refrigerted or were left t room temperture (with nd without the ddition of enzyl lohol), nd those of mllee slt gum (Eulyptus srgentii ssp. fllens) were used to test storge of frozen netr smples nd those to whih methnol hd een dded. Effet of wshing durtion To evlute the effet of longer wshing durtions on netr sugr olletion from individul flowers, three tretments were estlished. In eh tretment, sets of five flowers were removed from the plnt nd gitted in 2 ml of distilled wter for 1 min. The flowers in the first tretment were not FIG. 1. Eulyptus torqut (orl gum) flowers. Downloded from y guest on 4 My 218

3 Morrnt et l. Smpling nd storing florl netr t low volumes 535 soked; other flowers were soked for 19 min (seond tretment, totl 2 min) nd 59 min (third tretment, totl 6 min). Wshing durtions.1 min were hosen euse Käpylä (1978) found tht vrying wshing durtion etween.5 nd 1 min did not ffet the mss of sugrs removed from rinsed flowers. Smpling experiment To ompre the effets of netr olletion methods on netr sugr mss reovery, flowers of similr mturity (open nd showing no sign of senesene) were seleted from single plnt. Flowers from single plnt, nd of similr mturity, were seleted euse florl netr sugr rtios my vry etween individuls (Devlin et l., 1987; Hodges, 1993; Petnidou et l., 1996), nd with the ge of flowers (Petnidou et l., 1996). Netr ws smpled from these flowers using four ommonly employed methods: (1) miropillry tues (Collins nd Newlnd, 1986; MKenn nd Thomson, 1988; Kerns nd Inouye, 1993; Coret, 23); (2) filter pper wiks (MKenn nd Thomson, 1988; Kerns nd Inouye, 1993; Coret, 23); (3) wshing in known volume of distilled wter (Käpylä, 1978; Grünfeld et l., 1989); nd (4) rinsing with suessive rinses of known volume of distilled wter (Núñez, 1977; Mllik, 2). All smples were olleted onurrently, s netr sugr rtios my vry depending on olletion time (Petit nd Freemn, 1997). Netr volumes were too low for effetive extrtion using grvimetri tehniques, mirosyringes or miropipettes, nd these methods were not onsidered further. Miropillry tues. Seven flowers were smpled from the trget plnt using fixed-ore.25-ml miropillry tues (Drummond Sientifi Compny, Brooml, PA, USA). Netr ws drwn into the miropillry tues y mens of pillry tion. The volume of withdrwn netr ws then determined y mesuring the olumn of netr within the tue nd lulting the proportion of the entire olumn (.25 ml) filled. The netr in the miropillry tue ws then rinsed into 2 ml of distilled wter. Filter pper wiks. Seven flowers were smpled from the trget plnt using filter pper wiks. Equilterl tringles of Whtmn Numer 1 filter pper, with sides 1 m in length, were prepred. The filter pper ws held with foreps nd, in turn, eh orner ws pplied to the netries. Eh tringle of filter pper ws then pled in stoppered 3-mL vil ontining 2 ml of distilled wter, soked for 15 min, nd then gitted for pprox. 1 min until the pper ws redued to pulp. Wshing in distilled wter. The netr from eh of seven flowers ws smpled from the trget plnt using wshing method. Flowers were ut from the plnt, then eh flower ws pled in 3-mL vil nd 2 ml of distilled wter were dded using miropipette. The vil ws seled nd eh vil ws gitted mnully for 1 min. To determine how muh of the sugr ws removed in the initil wsh, eh flower ws then removed from the vil nd pled in fresh vil ontining 2 ml of wter. This proess ws repeted one dditionl time for totl of three wshes, nd the resulting solutions nlysed to determine wht proportion of the totl sugr ws olleted in eh wsh. For eh flower, the mount of eh sugr removed per wsh ws expressed s perentge of the totl sugr mss removed for ll wshes (Mllik, 2). The sugrs from these dditionl wshes were not, however, used for the smpling omprisons, s multiple wshes re not prtil in field sitution. Rinsing with known volume of distilled wter. Seven flowers were smpled from the trget plnt using rinsing method. A flower ws inverted over 3-mL vil nd four suessive rinses of distilled wter (.5 ml) were expelled over the florl netries using miropipette. This method did not require the removl of flowers from the plnt. It ws expeted tht rinsing ws the most urte smpling method euse, unlike the wshing method, sugrs potentilly leking from the ut surfe on the pediel would not e represented; it would lso e more likely to remove ll the netr from flower thn would other methods. Comprison of methods for the storge of netr solutions Storge methods for netr solutions were ompred over 28-d period. The im of this experiment ws to determine whether refrigertion nd/or the ddition of n ntimiroil gent, or freezing, would prolong the storge life of florl netr nd wter solutions. Ethnol ws not used s it interfered with the determintion of frutose using the HPLC method employed for sugr nlysis. Aetonitrile ws investigted in preliminry study, ut ws prone to ontmintion y ethnol. To prepre enough sugr solution for ll storge onditions, thes of 2 flowers from seven E. torqut were gitted in 5 ml (one flower per 2.5 ml) of wter for 1 min. All flowers were olleted within 1-h period t noon. For eh of four testing dtes (dys, 7, 14 nd 28), the solution ws divided into four vils, one for eh storge ondition. Sugr msses were not onverted to mg per flower for this experiment. Solutions were filtered into 8-mL rimp-top vils nd seled with 8-mm luminium silver sels with PTFE/silione liners (Gre Dvison Disovery Sienes, Bulkhm Hills, NSW, Austrli). The storge onditions, using E. torqut, were:() refrigertion t 4 ºC; () refrigertion t 4 ºC with the ntimiroil gent enzyl lohol (Hölzl, 23) (one enzyl lohol drop per ml of netr solution); () room temperture (pprox. 2 ºC); nd (d) room temperture with one drop of enzyl lohol per ml. An dditionl experiment exmined refrigertion with nd without methnol nd freezing, using the flowers of E. srgentii ssp. fllens. Beuse the flowers of this speies re smller thn those of E. torqut, 2 ml of wter nd 1 flowers were used (one flower per 2 ml). The storge onditions were: () refrigertion t 4 ºC (nlysed every dy for 7 d); () refrigertion t 4 ºC with methnol (1 ml of methnol per 1 L of netr solution; nlysed t, 7, 14 nd 28 d); nd () freezer 288C; nlysed t, 7, 14 nd 28 d). To mimi field olletion sitution, room temperture storge tretments were stored in non limte-ontrolled room, nd were sujet to mild flututions in mient temperture. Solutions were nlysed using HPLC on dy zero to provide ontrol smples, nd then 7, 14 nd 28 d lter. Only refrigertion with E. srgentii ssp. fllens ws exmined Downloded from y guest on 4 My 218

4 536 Morrnt et l. Smpling nd storing florl netr t low volumes every dy for week, exept dy 2, during whih n equipment prolem ws experiened. HPLC nlysis of netr solutions All solutions were filtered using.45-mm nylon memrne into 7 4 mm, 8-mL rimp-top vils, seled with 8-mm silver sels with PTFE/silione liners. Netr volumes of individul flowers were too low (,1 ml) to e nlysed with hnd-held sugr refrtometer. Sugr nlysis ws rried out y injeting 1 ml of filtered smple using n LS-32 utosmpler (SGE, Ringwood VIC, Austrli) nd hromtogrphed using BP-1 Crohydrte P þþ Column, mm (Benson Polymeri In., Reno, NV, USA) heted to 85 ºC using olumn heting module. A Wters 6E solvent delivery system ws used to deliver wter s the solvent t flow rte of.4ml min 21. Detetion ws hieved using Wters 41 refrtive index detetor heted to 4 ºC with sensitivity set t 256. Dt quisition nd hndling were done using Bseline 815 softwre from Wters. The msses of three netr sugrs (surose, gluose nd frutose) extrted from single flower were determined using this method. Dt nlysis The sttistil pkge SPSS w (Version 15.) ws used to ondut sttistil nlyses ( ¼.5). The unit of mesurement provided in the HPLC output ws in mg L 21 ; to onvert this onentrtion to sugr mss per flower, the output ws multiplied y.2 s 2 ml of distilled wter were used to wsh netr from flowers. Where more thn one flower ws used, the output ws divided y the numer of flowers (exept for storge, see ove). Sugrs olleted using eh smpling method nd wshing durtion were ompred using one-wy nlysis of vrine (ANOVA) fter ensuring tht the dt were normlly distriuted using Kolmogorov Smirnov goodness-of-fit test. Lest signifint differene men seprtion tehniques were used to identify the lotions of differenes etween smpling methods. Sugr onentrtions were exmined ross four test dtes for E. torqut (dys, 7, 14 nd 28) using repeted-mesures ANOVA (test dy su-smple from the solution mde with 2 flowers). In se of lk of spheriity, the Greenhouse Geisser orretion ws used; repeted ontrsts were used to ompre eh test dy ginst the previous one (F-tests for ontrsts re not presented). In ddition, x 2 ontingeny tle ws used to exmine ny differenes in rtios of sugr etween dy nd lter dtes (ounts onsisted of units of sugr in mg). To remin onservtive nd highlight even smll hnges in sugr mss, the experiment-wise lph level ws not djusted. Repeted-mesures ANOVA with repeted ontrsts (test dy su-smple from the solution mde with 1 flowers) ws used to exmine hnges in sugr onentrtion mong test dys:, 1, 3, 4, 5, 6 nd 7 in refrigerted smples without dditive. RESULTS Effet of wshing durtion Signifint differenes were found in the msses of gluose [F (2,12) ¼ 13.84, P ¼.1] nd frutose [F (2,12) ¼ 8.2, P ¼.6] mong wshing durtion tretments; however, msses of surose did not vry signifintly. Men seprtion tehniques showed no signifint differene etween the smples olleted fter 1 min of gittion in distilled wter nd those olleted fter 2 min of soking (Fig. 2). The tretment tht involved soking for 6 min ws signifintly different from the other two tretments with inresed msses of frutose (1 min, P ¼.3; nd 2 min, P ¼.6) nd gluose (1 min. P ¼.5; nd 2 min. P,.1). Smpling experiment Signifint differenes were deteted for the msses of ll sugrs mong the four smpling methods [surose, F (3,23) ¼ 7.877, P ¼.1; gluose F (3,23) ¼ 1.51, P,.1; frutose, F (3,23) ¼ 1.265, P,.1]. Men seprtion tehniques were pplied to determine the differenes in sugr yields etween pirs of tretments (Tle 1). The miropillry tue method olleted the lest sugr, yielding pproximtely four times less of ll three sugrs thn the two highest yielding methods (Fig. 3). The only exeption ws surose, for whih yields using filter pper tringles were not signifintly different. The filter pper tringle method yielded signifintly higher quntities of gluose nd frutose thn did the miropillry tues; it produed signifintly less surose thn did the wshing or rinsing methods, nd less frutose thn the rinsing method (Tle 1). Wshing removed slightly lower quntities of sugr thn did rinsing. The totl volume of netr sugr removed from E. spthult flower in eh wsh using three suessive wshes ws lso tested. The first two wshes removed pprox. 79% of the surose, 95% of the totl gluose nd 93% of the totl frutose (Tle 2). The proportion of the totl sugr mss removed from flowers in the first two wshes ws lower thn the 95% found y Mllik (2) for Euryphi luid. Sugr per flower (mg) Surose Gluose Frutose 1 min 2 min Durtion of wshing 6 min FIG. 2. Men sugr msses of surose, gluose nd frutose reovered from E. leuoxylon ssp. leuoxylon flowers, using three durtions of soking in distilled wter. The 1-min tretment ws gitted for 1 min; the 2-min tretment ws gitted for 1 min nd soked for 19 min; nd the 6-min tretment ws gitted for 1 min nd soked for 59 min. The error rs show the s.e. for eh sugr; n ¼ 5 for eh tretment. Different letters indite signifint differenes etween tretments for eh sugr onsidered seprtely. Downloded from y guest on 4 My 218

5 Morrnt et l. Smpling nd storing florl netr t low volumes 537 TABLE 1. Results of men seprtion testing (lest-signifint differene fter one-wy ANOVA) for the differenes in msses of surose, gluose nd frutose in E. spthult flowers mong four smpling methods Signifint differenes Sugr Tretment Miropillry tue Filter pper Wsh Rinse Surose Miropillry tue.1 ".1 " Filter pper.1 ".12 " Wsh.1 #.1# Rinse.1 #.12 # Gluose Miropillry tue.6 ",.1 ",.1 " Filter pper.6 # Wsh,.1 # Rinse,.1 # Frutose Miropillry tue.1 ".1 ",.1 " Filter pper.1 #.21 " Wsh.1 # Rinse,.1 #.21 # Vlues represent P-vlues nd indite signifint differenes etween tretments. Arrows must e red y olumn only nd indite the diretion of the differene (i.e. "¼ more, #¼ less). 6 5 Surose Gluose Frutose TABLE 2. Men perentges (+ s.e.) of sugrs removed from E. spthult flowers using three suessive wshes in 2 ml of distilled wter (n ¼ 7). Surose (%) Gluose (%) Frutose (%) Sugr (mg) Miropillry Filter pper Wsh Rinse Smpling method Storge of netr solutions F IG. 3. Men msses of the sugrs surose, gluose nd frutose in the netr reovered from the flowers of E. spthult using four smpling methods. Error rs show s.e.; n ¼ 7 for eh tretment. Different letters indite sttistilly signifint differenes etween tretments for eh sugr onsidered seprtely (see Tle 1 for detils). Note tht the P-vlue etween filter pper nd rinse for gluose ws.53. The six storge tretments (refrigertion, refrigertion with enzyl lohol or methnol, room temperture with nd without enzyl lohol, nd freezer) ll showed hnges in sugr omposition during the study period (Figs 4 nd 5). After 1 week, the mss of t lest one sugr hd hnged in ll tretments. Refrigerted or stored t room temperture, with nd without the ddition of enzyl lohol. In the tretments tht were First wsh Seond wsh Third wsh refrigerted or stored t room temperture, involving E. torqut, enzyl lohol did not prolong the life of ny of the smples. Although the F-tests were not signifint for ll three sugrs in the refrigertor enzyl lohol tretment euse of the lrge stndrd errors (on logrithmi sle, Fig. 4) nd lk of spheriity for frutose, implying the use of redued degrees of freedom, the men onentrtion of frutose t dy 14 ws.1 times tht t dy ( mg L 21 vs mg L 21, respetively), inditing very lrge vriility in mesurements for this tretment. At room temperture with enzyl lohol, the F-tests for ll sugr onentrtions were signifintly different, with the sme pttern of lrge mens nd stndrd errors for frutose t dy 14 nd dy 28 ompred with dy (Fig. 4). At dy 28, the smples ontining enzyl lohol showed gretly inresed sugr levels, prtiulrly in the se of frutose, on the HPLC hromtogrms. For exmple, t dy 28, the men frutose mss of the refrigerted smples tht hd een treted with enzyl lohol ws.4 times greter thn those of the sme tretment t dy 14 (Fig. 4). Refrigertion without enzyl lohol produed the est results in this set of experiments; however, the onentrtion of frutose hd hnged y dy 7. Confidene intervls t dy for omined smples in the refrigertor (smples not yet refrigerted) nd room temperture tretments were reltively lrge (surose ¼ mg L 21, gluose ¼ mg L 21 nd frutose ¼ mg L 21 ), showing pereptile error in mesurement when sugr onentrtions re high. Downloded from y guest on 4 My 218

6 538 Morrnt et l. Smpling nd storing florl netr t low volumes Sugr (mg L 1 ) A Surose Gluose Frutose A Surose Gluose Frutose d Sugr (mg L 1 ) Sugr (mg L 1 ) Sugr (mg L 1 ) B C D Dy Dy 7 Dy 14 Dy 28 Storge durtion Sugr (mg L 1 ) Sugr (mg L 1 ) B Dy Dy 7 Dy 14 Dy 28 Storge durtion d FIG. 4. Men onentrtion of surose, gluose nd frutose in wsh solutions of E. torqut for four netr storge onditions over 28-d period. Error rs show s.e. Sugr onentrtion in mg L 21 is expressed on logrithmi sle. (A) Refrigertor, (B) refrigertor with enzyl lohol, (C) room temperture, (D) room temperture with enzyl lohol (note different vertil sle to A C). Different letters indite sttistilly signifint differenes etween times of storge for eh sugr onsidered seprtely. Frozen or treted with methnol. Using E. srgentii ssp. fllens, refrigertion with methnol nd freezing t dys, 7, 14 nd 28 ws lso tested, nd we found tht gluose ws stle in the methnol tretment, ut surose nd frutose hd F IG. 5. Men onentrtion (mg L 21 ) of surose, gluose nd frutose in wsh solutions from flowers of E. srgentii ssp. fllens for two storge onditions over 28-d period. Error rs show s.e. (stndrd errors for surose were generlly too smll to e visile t this sle). Sugr in mg is expressed on logrithmi sle. (A) Freezer, (B) refrigertor with methnol. Different letters indite sttistilly signifint differenes etween times of storge for eh sugr onsidered seprtely. Note tht no dt were reorded for surose in the refrigertor with methnol tretment on dy 14. hnged y dy 7 (F ¼ , d.f. ¼ 2, P,.1 for the ANOVA; Fig. 5). Surose nd frutose onentrtions hd hnged y dy 7 in the freezer tretment (F ¼ , Downloded from y guest on 4 My 218

7 Morrnt et l. Smpling nd storing florl netr t low volumes 539 d.f. ¼ 3, P,.1; F ¼ 72.28, d.f. ¼ 3, P,.1, respetively, for the ANOVA); gluose hd hnged etween dy 7 nd dy 14 (F ¼ 8.651, d.f. ¼ 3, P ¼.1 for the ANOVA). Overll it ppers tht refrigertion lone produes the est results in terms of sugr onentrtion, with reltively stle onentrtions of gluose nd surose for t lest 2 weeks; surose nd frutose hd hnged y dy 7 in the freezer nd refrigertor þ methnol tretments. Further experimenttion t finer sle with E. srgentii ssp. fllens to detil hnges in sugr onentrtion within 1 week using refrigerted smples without dditive showed tht the gluose nd frutose onentrtions hd hnged y dy 3 (F ¼ 4.638, d.f. ¼ 6, P ¼.1; F ¼ 21.33; d.f. ¼ 2.625, P,.1, respetively); however, the onentrtion of these sugrs did not follow desending or sending trend nd n error in mesurement my hve een responsile for the differenes found (Fig. 6). In ft, the onfidene intervls tht were lulted for dy for omined smples did not indite lrge error in mesurement (surose ¼ mg L 21 ; gluose ¼ mg L 21 ; frutose ¼ mg L 21 ). The differenes found etween dys in the 7-d tril for gluose nd frutose, lthough sttistilly signifint, my not hve iologil relevne. x 2 nlysis for tretments involving four storge durtions reveled tht t dy 7 the room temperture solution to Sugr (mg L 1 ) Sugr (mg L 1 ) Sugr (mg L 1 ) Surose Gluose Frutose Dy Dy 1 Dy 3 Dy 4 Dy 5 Dy 6 Dy 7 Dy of nlysis F IG. 6. Men onentrtion (mg L 21 ) of surose, gluose nd frutose in refrigerted netr smples from E. srgentii ssp. fllens, with no dditive, over 7-d period. Error rs show s.e.; n ¼ 7 for eh tretment. Gluose nd frutose onentrtions hd hnged signifintly y dy 3. Different letters represent signifint differenes etween djent rs only. whih enzyl lohol hd een dded showed signifint hnges in sugr rtios when ompred with dy (x 2 ¼ 8.42, d.f. ¼ 2, P,.5). The differene in sugr rtios etween dy nd dy 14 ws strongly signifint for the refrigerted solution to whih enzyl lohol ws dded (x 2 ¼ 94.85, d.f. ¼ 2, P,.1). The sugr rtios in the freezer tretment hd not hnged y dy 28, nor hd those in the refrigertor þ methnol tretment; the remining tretments experiened signifint hnge y dy 28 (refrigertor with no enzyl lohol, x 2 ¼ 13.58, d.f. ¼ 2, P,.1; room temperture with no enzyl lohol, x 2 ¼ 37.4, d.f. ¼ 2, P,.1). Until dy 28, the sugr rtios re thus representtive of those in netr on the olletion dy when using the freezer or refrigertor þ methnol storge tretment. DISCUSSION Signifint vritions were found in the effetiveness of the vrious methods for smpling nd storing florl netr from flowers with low netr volumes. The results for florl netr nlysis show tht: () it should e olleted using wshing (1 min) or rinsing method (ut only 6% of surose is reovered in the first wsh, sed on three wshes, ompred with 9% gluose nd 82% frutose); () it should e refrigerted immeditely fter olletion with no dditive if sugr onentrtion is to e exmined; () it should e nlysed immeditely for sugr mss if ury is needed; nd (d) the sugr rtios remin stle for t lest 14 d in the refrigertor nd 28 d in the freezer nd refrigertor þ methnol tretments. Netr smpling The 1-min gittion method ws deemed to e the most suitle method for smpling in the field euse it is more time effiient thn is the 2-min tretment. The wsh durtion experiment showed tht smples olleted using this method re potentilly prone to ontmintion y extrflorl sugrs if soked or gitted for too long. It is possile tht the onsiderle inrese in sugrs in the 6-min wshing tretment is result of vsulr nd ellulr fluid lehing from the ut surfe on the pediel during soking, s ws found y Herrer et l. (26) in their investigtion of the netr sugr omposition of the perennil her Helleorus foetidus (Rnunulee). Seling the ut surfe with wx (D. W. Inouye, University of Mrylnd, pers. omm.) or surgil glue my ovite this prolem. Conversely, the method my lso remove more of the netr sugrs thn do other methods, or netries my lek sugrs euse of osmoti stress. The first two 1-min wshes of flower were estimted to remove pprox. 8% of the totl sugrs ut, if lehing from the pediel ourred, this numer would overestimte the totl mount of netr sugr removed. All smpling methods re likely to underestimte gretly the mounts of sugrs present in flowers sine one of the est methods only ollets 6 9% of the sugrs in the first wsh, s estimted y three wshes. This point needs to e tken into ount for studies on energetis, for exmple. It ws lso noted tht proportions of sugrs extrted in eh wsh differ; surose lehing from the ells or phloem (Pte et l., 1985; De l Brrer nd Noel, 24) my ontriute Downloded from y guest on 4 My 218

8 54 Morrnt et l. Smpling nd storing florl netr t low volumes to reltively high perentge of this sugr reovered fter the first wsh. Surose is inverted to gluose nd frutose in the netries y invertse (Pte et l., 1985). The rinsing nd wshing methods pper to e the most effetive wys of determining the mss of sugrs in the netr of smll flowers, lthough using distilled wter my use lehing of sugr from the flower s ells s result of osmoti pressure. Wshing or rinsing liquids simulting the sliv (t lest its ioni strength nd ph) of flower visitors my e more pproprite to understnd how muh sugr is ville to them nd their impt on netr prodution. These onsidertions side, the rinsing method yielded the highest msses of sugrs, nd, euse it does not require flowers to e removed from the plnt, it n e used for studies of netr replenishment where repeted mesurements re required. Also, this method does not introdue non-florl sugrs into the netr smples. If we ssume tht rinsing is the most urte smpling method euse it removes the most netr sugr, nd is less likely to e ontminted with extrflorl sugrs, wshing, whih is not signifintly different, is the most prtil field method. The wshing method is quik, esy method of estimting florl sugr in flowers with low netr volumes nd is not limited y netr sugr onentrtion or flower struture. It is lso less time onsuming thn is the rinsing method, nd requires less equipment. Although it does not remove ll sugr from the flower in the first extrtion, it is the simplest nd most effetive method of quntifying netr sugrs in situtions where smll flowers, low florl netr volumes or omplex strutures hinder the olletion of netr using other methods. This method is reommended where repeted mesurements of netr re not needed. Blotting up netr with filter pper wik nd miropillry extrtion methods were generlly less effetive methods of removing ll sugrs from flower thn were the rinsing nd wshing methods. Although removl of florl netr using miropillry tues n provide n estimte of the netr volume of single flower ( ml for E. spthult, n ¼ 7), the results of the HPLC nlysis show tht this method is likely to underestimte florl netr volume gretly. It n e diffiult to ensure tht ll netr hs een removed from flowers using filter pper or miropillry tues, prtiulrly in speies tht hve smll flowers, low florl netr volumes or omplex strutures. The filter pper nd miropillry methods require delite movements, nd mke extrtion of netr from the flowers of suh speies diffiult. These prolems re ompounded in diffiult field onditions suh s drkness or extreme wether ( pers. os.). Although some studies hve found tht miropillry tues nd filter pper wiks re urte methods for extrting netr from flowers with low netr volumes (MKenn nd Thomson, 1988), the results of this study indite tht miropillry tues re less effetive thn re other methods for eulypts. In flowers with low netr volumes, volume is diffiult to estimte. Although miropillry tues re not s effetive for sugr reovery s the other methods of smpling florl netr sugrs tested in this study, they my prove useful s tools to estimte netr volume per flower. If sugr onentrtion olleted on verge y miropillry tue is known nd we ssume tht this onentrtion represents the flower netr, then totl flower netr volume n e extrpolted from sugr mss. For exmple, the men netr volume olleted per flower using miropillry tues ws.18 ml (+.16) for E. spthult, nd the men totl sugr mss ws.17 mg (+.19; n ¼ 7). The men totl sugr mss reovered with the rinse method ws 1.7 mg (+.18; n ¼ 7), so the men volume per flower is estimted to e: (miropillry netr volume sugr mss in rinse)/sugr mss in miropillry tue, or 1.15 ml,.6 times more thn estimted with the miropillry tue lone. Netr storge The results of the storge trils highlight the need to nlyse netr smples in timely fshion. While the onentrtion of t lest one sugr ws signifintly different y dy 7 in ll tretments, the rtios of sugrs did not vry signifintly efore tht dy, nd hd not hnged t dy 28 in the freezer nd refrigertor þ methnol storge tretments, nd dy 14 in the refrigertor tretment. Thus, netr solutions prepred for the investigtion of sugr rtios my e stored for periods of t lest 28 d if they re frozen, or refrigerted nd treted with methnol, or 14 d if they re refrigerted with no dditive. For the exmintion of sugr onentrtions, it is preferle to refrigerte smples with no dditive. Over 7-d tril period, sttistilly signifint ut numerilly minor flututions were oserved, whih did not follow trend. At orser (weekly) sle of nlysis, no differene ws found in oth surose nd gluose onentrtions etween dy nd dy 7 in the refrigerted smples, wheres other tretments did not perform s well (refrigertion with enzyl lohol, lthough performing stisftorily until dy 7, ws ndoned euse of the lrge impt it hd on frutose redings fter tht time). Storge onditions should e hosen depending on the ojetives of the study. If very urte sugr onentrtions re needed, then immedite nlysis of the smples is reommended; otherwise, smples my e stored for 7 d. Frutose onentrtion mesurements tend to hnge more drmtilly thn those of surose nd gluose, so rpid proessing is impertive where ext mesurements of this sugr re needed. It should e noted, however, tht mesurements my vry depending on the method of nlysis. The dt shown in Fig. 4 indite tht the onentrtion of frutose in the smple hd inresed signifintly fter storge for 14 d, or longer, when enzyl lohol hd een dded s preservtive. The only likely soure of frutose is from the hydrolysis of surose, so suh mrked inreses in msses re unlikely. Sine the mss of surose in the smples is very smll, hydrolysis of surose to produe frutose nd gluose nnot ount for the drsti inrese in frutose fter only 14 d of storge. It ppered tht the ddition of enzyl lohol might e the use of this inrese in frutose. Oxidtion of enzyl lohol during storge my led to formtion of produts suh s enzldehyde nd/or enzoi id; these ompounds my o-elute with frutose during hromtogrphi nlysis, resulting in nomlously high frutose onentrtion. To test this hypothesis, queous solutions of enzldehyde nd enzoi id were hromtogrphed under Downloded from y guest on 4 My 218

9 Morrnt et l. Smpling nd storing florl netr t low volumes 541 the sme onditions s those used for sugr nlysis. The results of this experiment demonstrted tht enzldehyde elutes t retention time very lose to tht of frutose. Sine the mount of enzldehyde, whih must hve resulted from oxidtion of the enzyl lohol dded to the smples, is likely to e present in muh higher onentrtion thn frutose, it is prole tht the frutose pek ws missed nd tht the enzldehyde pek ws mesured s frutose. This misinterprettion will hve led to the pereived inreses in the mount of frutose present in the storge smples to whih enzyl lohol ws dded. Chnges in the sugrs re diffiult to explin; miroil tivity n rete rel hnges in sugrs nd/or rete other ompounds tht interfere in the ssy, resulting in pprent hnges tht my not e rel. If it is the se, pulsed mperometri detetion (PAD) might overome prolems of pprent hnges in sugr msses due to interferenes, euse of the speifiity of this method s opposed to refrtive index detetion. The results of netr nlysis n vry onsiderly depending on the smpling nd storge methods hosen. To smple netr from plnts with low florl netr volume in field sitution, it ppers tht the 1-min wsh method is the most effiient, nd tht the smple should e refrigerted without dditive. The smples should e nlysed s soon s possile upon return to the lortory, nd preferly within dy. ACKNOWLEDGEMENTS We thnk I. Ametov for her ssistne in the lortory, J. Sloome, R. Aei, D. Crver, nd H. West for their help with otining equipment, nd P. Rismiller, R. Shrrd, W. Sowroft, D. Inouye, M. Sedgley nd n nonymous reviewer for reviewing the mnusript; the ltter inspired the osmoti sttement. This pper ws prt of D.M. s Honours thesis in the progrmme of Applied Siene (Hons) (Biodiversity, Environmentl nd Prk Mngement). This work ws supported y the Ntionl Geogrphi Soiety (87-6 to S.P., with A. Shrp, nd the ollortion of D.M. nd R.S.); the Nture Foundtion of South Austrli (grnt to S.P.); the Field Nturlists Soiety of South Austrli (Lirend Endowment Fund Reserh Grnt to D.M.); the Deprtment for Environment nd Heritge (South Austrli; Northern nd Yorke Region); the University of South Austrli (UniSA) (trvel); nd the UniSA Division of Informtion Tehnology, Engineering nd the Environment (Seed Funding to S.P. nd R.S.; Summer Sholrship to Promising Undergrdute Student D.M.). LITERATURE CITED Armstrong DP, Pton DC Methods for mesuring mounts of energy ville from Bnksi infloresenes. Austrl Eology 15: Birthnell MJ, Gison M. 26. Long-term flowering ptterns of melliferous Eulyptus (Myrtee) speies. Austrlin Journl of Botny 54: Bolten AB, Feinsinger P Why do hummingird flowers serete dilute netr? Biotropi 1: Bolten AB, Feinsinger P, Bker HG, Bker I On the lultion of sugr onentrtion in flower netr. Oeologi 41: Bond HW, Brown WL The exploittion of florl netr in Eulyptus inrsst y honeyeters nd honeyees. Oeologi 44: Collins BG, Newlnd C Honeyeter popultion hnges in reltion to food vilility in the jrrh forest of Western Austrli. Austrlin Journl of Eology 11: Coret SA. 23. Netr sugr ontent: estimting stnding rop nd seretion rte in the field. Apidologie 34: 1 1. De l Brrer E, Noel PS. 24. Netr: properties, florl spets, nd speultions on origin. Trends in Plnt Siene 9: Devlin B, Horton JB, Stephenson AG Ptterns of netr prodution of Loeli rdinlis. Amerin Midlnd Nturlist 117: Dungn RJ, Beggs JR, Wrdle DA. 24. A simple grvimetri tehnique for estimting honeydew or netr prodution. New Zelnd Journl of Eology 69: Erhrdt A, Bker I Pollen mino ids n dditionl diet for netr-feeding utterfly? Plnt Systemtis nd Evolution 169: Grünfeld E, Vinent C, Bgnr D High performne liquid hromtogrphy nlysis of netr nd pollen of strwerry flowers. Journl of Agriulturl nd Food Chemistry 37: Hinsworth FR, Wolf LL Netr hrteristis nd food seletion y hummingirds. Oeologi 25: Heinrih B Energetis of pollintion. Annul Review of Eology nd Systemtis 6: Herert SM, Clder WA Sodium, potssium, nd hloride in florl netrs: energy-free ontriutors to refrtive index nd slt lne. Eology 64: Herrer C, Pérez R, Alonso C. 26. Extreme intrplnt vrition in netr sugr omposition in n inset-pollinted perrenil her. Amerin Journl of Botny 93: Hodges SA Consistent interplnt vrition in netr hrteristis of Mirilis multiflor. Eology 74: Hölzl W, Koppold J, Mrquis-Bienwld S, Preuss A. 23. Benzyl lohol derivtives nd their use s ntimiroil gents. Online ptent: jsp?wo= &ia=ep232618&display=status Inouye DW, Fvre ND, Lnum JA, et l The effets of nonsugr netr onstituents on estimtes of netr energy ontent. Eology 61: Käpylä M Amount nd type of netr sugr in some wild flowers in Finlnd. Annles Botnii Fennii 15: Kerns CA, Inouye DW Tehniques for pollintion iologists. Colordo: University Press of Colordo. Lnz J, Smith GC, Sk S, Csh A Vrition in netr volume nd omposition of Imptiens pensis t the individul, plnt, nd popultion levels. Oeologi 12: Lloyd S, Ayre DJ, Wheln RJ. 22. A rpid nd urte visul ssessment of netr prodution n revel ptterns of temporl vrition in Bnksi ereifoli (Proteee). Austrlin Journl of Botny 5: Mllik SA. 2. Tehnique for wshing netr from the flowers of Tsmnin letherwood (Euryphi luid: Euryphiee). Austrl Eology 25: Mnets Y, Petropoulou Y. 2. Netr mount, pollintor visit durtion nd pollintor suess in the Mediterrnen shru Cistus retius. Annls of Botny 86: Mrtínez del Rio C, Bker HG, Bker I Eologil nd evolutionry implitions of digestive proesses: ird preferenes nd the sugr onstituents of florl netr nd fruit pulp. Cellulr nd Moleulr Life Sienes 48: MKenn MA, Thomson JD A tehnique for smpling nd mesuring smll mounts of florl netr. Eology 69: Núñez JA Netr flow y melliferous flor nd gthering flow in Apis mellifer ligust. Journl of Inset Physiology 23: Pte JS, Peoples MB, Storer PJ, Atkins CA The extrflorl netries of owpe (Vign unguiult (L.) Wlp.). II. Netr omposition, origin of netr solutes, nd netry funtioning. Plnt 166: Petnidou T, Vn Lere AJ, Smets E Chnge in florl netr omponents from fresh to senesent flowers of Cppris spinos (Cppridee), noturnlly flowering Meditternen shru. Plnt Systemtis nd Evolution 199: Petit S, Freemn CE Netr prodution of two symptri speies of olumnr ti. Biotropi 29: Petit S, Pors L Survey of olumnr ti nd rrying pity for netr-feeding ts on Curço. Conservtion Biology 1: Downloded from y guest on 4 My 218

10 542 Morrnt et l. Smpling nd storing florl netr t low volumes Shondue JE, Mrtínez del Rio C. 23. Conentrtion-dependent sugr preferenes in netr-feeding irds: mehnisms nd onsequenes. Funtionl Eology 17: Stephnou M, Petropoulou Y, Georgiou O, Mnets Y. 2. Enhned UV-B rdition, flower ttriutes nd pollintor ehviour in Cistus retius: Mediterrnen field study. Plnt Eology 147: Swnson CA, Shuel RW The entrifuge method for mesuring netr yield. Plnt Physiology 25: Tshpk M. 24. Energy density ptterns of netr resoures permit oexistene within guild of Neotropil flower-visiting ts. Journl of Zoology 263: Willmer P The effets of inset visitors on netr onstituents in temperte plnts. Oeologi 47: Downloded from y guest on 4 My 218