COMPARISON OF DIFFERENT METHODS TO MEASURE THE INFESTATION LEVEL OF WINTERING HONEYBEE COLONIES BY THE SPORES OF NOSEMA APIS Z.

tekst: Krystyna Pohorecka, Janina Muszyńska

Pszczelnicze Zeszyty Naukowe, Rok XL, 1996 — Krystyna Pohorecka, Janina Muszyńska

strona 129

PSZCZELNICZE ZESZYTY NAUKOWE Rok XL, Nr 1 1996

Krystyna Pohorecka, Janina Muszyńska

Institute of Pomology and Floriculture, Bee Division, ul. Kazimierska 2, 24-100 Puławy, POLAND

Summary

There is a wide diversity of methods used to assess the extent of infestation of honeybee colonies by Nosema apis. The diversity lies not only in different sampling techniques but also in the microscopic examination methods and, consequently, in the manner in which the results are presented.

The objective of the study was to compare two methods for the determination of the level of infestation by Nosema apis of honeybee colonies: one based on the mean number of spores per 1 bee and the other based on the percentage of affected bees.

Suitability of two kinds of bee samples used to measure Nosema spore infestation of wintering bees i.e sampling live bees collected from the cluster vs. sampling dead bees collected from the winter fall was also compared.

The results from the study indicate that the method which measures the infestation of bee colonies by the mean number of Nosema spores per 1 bee is just as accurate as the one involving the estimation of the percentage of affected bees in a sample of 10 individuals. The results also bear out earlier reports indicating a smaller reliability of dead bees as compared to the living ones as the measurre of Nosema disease of bee colonies.

Keywords: Nosema apis, infestation level, diagnostics, method.

INTRODUCTION

The protozoan Nosema apis responsible for the nosema disease of adult honeybees (Apis mellifera) parasitizes and multiplies in the cytoplasm of the epithelium cells of the intestine. In the course of the parasite’s development cycle new infestation forms – spores – are generated.

The diagnostics of Nosema infestation in honeybee colonies is based on the detection – by laboratory methods – of the presence of spores in the tissues and in the alimentary tract of worker bees. There is a wide diversity of methods used to make the determinations of Nosema infestation level of honeybee colonies. The diversity consists not only in the sampling approaches but also in the methods employed for microscopic examinations and, consequently, in the manner in which the results are presented.

Both living and dead bees are taken for the examinations involving

strona 130

from 10 to 100 individuals. Bees from a given sample may be examined individually or jointly and thus make up a composite sample.

Individual examinations consist in grinding and suspending in water single abdomens or dissected alimentary tracts of worker bees. Drops of the suspensions are examined under the microscope for the presence of spores. The result is expressed as the percentage of infested bees in the sample (Bailey 1953, 1967; Doul and Celier 1961; Gross and Ruttner 1970; Laere et al. 1980). The result can also be expressed as the mean number of spores per 1 bee in a sample from a given colony provided spores in individual preparations are counted (Gross and Ruttner 1970, Laere et al. 1980, Pickard and El-Shemy 1989).

In bulked examinations, a homogenized suspension of all bees of the sample is prepared to make a microscopic preparation. The estimation of infestation level is presented as the mean number of spores per 1 bee in a sample collected from a colony (Gochnauer 1957, Doul and Eckert 1962, L’Arrivee 1963a, Cantwell 1970).

The results of comparison studies on the suitability and accuracy of the individual method vs. the bulked method are equivocal.

Doul (1965) found that the number of spores determined in small bee samples failed to correlate with infestation expressed as percentage. Fries et al. (1984) found that individual examinations performed on 60 bees, although more labour-consuming, reflect better the actual infestation level that the bulked examinations. Pickard and El-Shemy (1989) in their investigations of the seasonal changes in Nosema disease of honeybee colonies used individual bee examinations within samples comprising 30 individuals each and arrived at a conclusion that infestation level expressed as the percentage of affected bees is an estimation accurate enough.

However, L’Arrivae showed that a joint examination of 25 bees is just as reliable as an individual analysis of 10 bees. His data failed to reveal any significant differences in the accuracy of the methods under comparison. Likewise, Bailey (1968) found that the number of spores in samples of 25 bees each was positively correlated with the mean number of individually examined bees which were found to be Nosema-infested. Similar results were obtained by Fingler (1982). According to that investigator the correlation coefficient between the mean number of spores per bee in bulked samples and the percentage of infested bees is highly significant (r=0.859) which indicates that the calculated number of spores per bee is related to the number of infested bees in the sample. Fries et al. (1984) compared three methods for the estimation of infestation level of bee colonies by Nosema in the spring and, at the same time, evaluated the suitability of those methods for the assessment of the effect of Nosema infection on honey production. Both dead and living bees were collected by

strona 131

those investigators to be examined both individually and in bulk. The comparison of the mean number of spores per 1 living bee with the mean number of spores per 1 dead bee revealed a significant positive correlation but the coefficient was lower than 50%. The highest correlation between honey output and infestation level became manifest during the determination of infestation level made on living bees. The investigators suggest that in the spring the evaluation of the effect of the infestation by the protozoan N.apis in bee colonies ought to be made by examining living bees. According to them the results obtained using that method are more accurate and better reflect the actual health status of the colonies.

The investigations presented in the study had two principal objectives:

  1. The evaluation of the relationship between two methods for the estimation of the infestation of wintering bee colonies by N.apis spores
  2. method 1 – based on the mean number of spores per bee
  3. method 2 – based on the percentage of infested bees in the sample
  4. The evaluation of the relationship between two kinds of bee samples used to estimate the extent of infestation by N.apis spores of bee colonies
  5. living bees (from the cluster)
  6. dead bees

MATERIAL AND METHODS

In this study we made use of the data which were obtained in the years 1978-1981 in the course of the investigation on the physiology of wintering of honeybee colonies (Muszyńska, Bornus 1983a, 1983b). The investigations were conducted in Puławy, in the apiary of the Bee Division of the Institute of Pomology and Floriculture. The colonies to be examined had queens of the Caucasian race which had been naturally inseminated and established in Dadant type bee hives. The successive wintering seasons in the years 1978-1981 differed considerably both for the duration of wintering and for the length of the transition period. Likewise, the last autumn flights and the first spring flights also took place on different dates.

In each of the three years six honeybee colonies were included in the investigation. Samples to be examined for Nosema infestation were taken over two successive years from November to April and in the third year from November to March at monthly intervals. On each sampling date two kinds of bees were collected for examination:

        living – collected from the top of the cluster without disturbing it         dead – fallen during 7 days prior to sampling date.

strona 132

There were 10 bees in each sample.

The evaluation of the infestation level in a sample was made based on the examination of individual worker bees. The abdomen of each bee was ground in 5 ml of distilled water. The number of spores present in the homogenized suspension was counted in each of 10 large fields of Bürker’s haemacytometer. The number of spores per 1 bee was calculated according to the formula:

total number of spores from all fields ————————————————————————————————————— x 4 x 10^6 x 5 number of fields

The results for each of the 10 bees were added and subsequently the mean number of spores per one bee in a sample was calculated (method 1). The percentage of infested bees in the sample was also calculated (method 2). A total of 192 samples, each comprising 10 bees, were examined which makes up a grand total of 1920 bees.

The results were subjected to statistical analysis.

Correlation was studied between the data on infestation level obtained by using methods 1 and 2. Interrelationships between data on infestation level obtained from the examination of living and dead bees collected simultaneously from the same colonies was also estimated. Regression analysis was used to make the estimation. The calculations were made using the mean values of spores per one bee and of percentage of infested bees.

Comparisons were made for: a) results obtained from 6 experimental colonies over the total of three successive years, b) results obtained separately for each colony and for each year, c)results obtained for the same months totalled over the three successive years.

A total of 40 regression analyses were performed.

RESULTS

The results of correlation between the infestation of bee colonies, expressed as the mean number of spores per bee, and the percentage of infested bees in the samples are listed in Tables 1a, 1b, 1c.

strona 133

Correlation coefficients (r) and determination coefficients (D) between different infection levels by N.apis in bee colonies estimated two methods Współczynniki korelacji (r) i współczynniki determinacji (D) pomiędzy różnymi poziomami infekcji N.apis w rodzinach pszczelich

X1 = mean number of spores per living bee – średnia liczba spor/pszczołę żywą Y1 = % infested bees in a sample of living bees – % porażonych robotnic w próbie żywych pszczół X2 = mean number of spores per dead bee – średnia liczba spor/pszczołę martwą Y2 = % of infested bees in a sample of dead bees – % porażonych robotnic w próbie pszczół martwych n = number of replications – liczba powtórzeń

T a b l e 1a

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badane cechy n r D (%)
Mean X and Y values over the total of X1 on Y1 15 0.818** 67.03
three years X2 on Y2 15 0.943** 88.95
Wartości średnie X i Y – 3 lata łącznie

T a b l e 1b

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badane cechy n r D (%)
data from 1978/79 wintering X1 on Y1 36 0.705** 49.77
dane z zimowli 1978/79 X2 on Y2 36 0.551** 30.41
data from 1979/80 wintering X1 on Y1 36 0.730** 53.42
dane z zimowli 1979/80 X2 on Y2 36 0.725** 52.69
data from 1980/81 wintering X1 on Y1 36 0.840** 70.58
dane z zimowli 1980/81 X2 on Y2 36 0.943** 71.86

T a b l e 1c

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badane cechy n r D (%)
November – Listopad X1 on Y1 12 0.921** 84.97
X2 on Y2 12 0.850** 72.25
December – Grudzień X1 on Y1 18 0.553** 30.61
X2 on Y2 18 0.896** 80.34
January – Styczeń X1 on Y1 18 0.935** 87.52
X2 on Y2 18 0.961** 92.46
February – Luty X1 on Y1 18 0.970** 94.10
X2 on Y2 18 0.828** 68.71
March – Marzec X1 on Y1 12 0.912** 83.28
X2 on Y2 12 0.785** 61.65
April – Kwiecień X1 on Y1 12 0.811** 65.80
X2 on Y2 12 0.675* 54.35
  • significant correlation p=0.05 korelacja istotna p = 0,05

** highly significant correlation p=0.01 korelacja wysoko istotna p = 0,01

strona 134

All the analyses showed the occurrence of positive, highly significant correlations between the data on the infestation by spores of N.apis of honeybee colonies, expressed as the mean number of spores per bee, and the data on infestation expressed as the percentage of infested bees in the sample. Those relationships held true of the evaluation of data which related both to the samples of living bees collected from the cluster and to dead bees from the winter fall.

The results suggest that the mean number of Nosema apis per 1 bee in a sample of 10 individuals is closely related to the percentage of infested bees in a given sample. The relationship occurs irrespective of the kind of bees collected for the examinations.

The results on correlation between the mean number of spores per living bee and the mean number of spores per dead bee and on correlation between the percentage of infested living bees and that of dead bees in bee samples collected simultaneously from the same colonies are shown in Tables 2a, 2b, 2c.

Correlation coefficients (r) and determination coefficients (D) between different levels of infestation by N.apis in bee colonies as determined on living bees and dead bees Współczynniki korelacji (r) i współczynniki determinacji (D) pomiędzy różnymi poziomami infekcji N.apis w rodzinach pszczelich oznaczonymi na pszczołach żywych i martwych

X1 = mean number of spores per living bee – średnia liczba spor/pszczołę żywą Y1 = % infested bees in a sample of living bees – % porażonych robotnic w próbach pszczół żywych X2 = mean number of spores per dead bee – średnia liczba spor/pszczołę martwą Y2 = % of infested bees in a sample of dead bees – % porażonych robotnic w próbach pszczół martwych n = number of replications – liczba powtórzeń

T a b l e 2a

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badana cecha n r D (%)
Mean X and Y values over the total of X1 on Y1 15 0.850** 72.32
three years X2 on Y2 15 0.588* 34.58
Wartości średnie X i Y – 3 lata łącznie

T a b l e 2b

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badana cecha n r D (%)
data from 1978/79 wintering X1 on Y1 36 0.102 1.04
dane z zimowli 1978/79 X2 on Y2 36 0.444** 19.71
data from 1979/80 wintering X1 on Y1 36 0.350* 12.28
dane z zimowli 1979/80 X2 on Y2 36 0.354* 12.54
data from 1980/81 wintering X1 on Y1 18 0.281* 7.94
dane z zimowli 1980/81 X2 on Y2 18 0.452* 20.46
strona 135

T a b l e 2c

Type of data for regression analysis / Dane do obliczenia analizy regresji Trait / Badana cecha n r D (%)
November – Listopad X1 on Y1 12 0.471 22.19
X2 on Y2 12 0.239 5.74
December – Grudzień X1 on Y1 18 0.448* 20.09
X2 on Y2 18 0.558* 31.16
January – Styczeń X1 on Y1 18 0.737** 54.41
X2 on Y2 18 0.689** 47.50
February – Luty X1 on Y1 18 0.363 13.22
X2 on Y2 18 0.489* 23.95
March – Marzec X1 on Y1 12 0.036 0.13
X2 on Y2 12 0.295 8.71
April – Kwiecień X1 on Y1 12 0.022 0.05
X2 on Y2 12 0.391 15.35
  • significant correlation p=0.05 korelacja istotna p = 0,05

** highly significant correlation p=0.01 korelacja wysoko istotna p = 0,01

The comparison of the mean number of spores per 1 living bee with the mean number of spores per one dead bee in samples collected from the same colonies showed the absence of correlation between the respective values in the majority of analyses. It was only for the comparison of mean values totalled over three years that a significant, positive correlation was found (Table 2a). There was another significant correlation in 1979/80 but the coefficient „r” and determination coefficient were low (Table 2b). Over the three successive years the traits under investigations showed correlation only for December (a low determination coefficient) and in January (Table 2c).

The analysis based on the comparison of the percentage of infested worker bees in the cluster with the percentage of worker bees in winter fall samples also failed to yield unambiguous results. Even though the three-year means of the investigated traits and the values for individual years proved to be correlated the interrelations were characterized by low correlation coefficients and low determination coefficients (Tables 2a, 2b). Mutual relationships between the traits for individual months over the three successive years of the study were also in this case found only for December and January (Table 2c).

The results give evidence that significant differences exist between the infestation level of bee colonies by spores of N.apis as calculated from the number of living bees collected from the cluster and the infestation level obtained from examining samples of dead bees from the winter fall. The

strona 136

differences occurred irrespective of the evaluation method used and the level of infection was always higher in the samples of dead bees.

DISCUSSION AND CONCLUSIONS

The results of the study indicate that the method to estimate the level of infestation of honeybee colonies by spores of N.apis based on the mean number of spores per 1 bee is just as accurate as the method consisting in the calculation of the percentage of dead bees in a 10-bee sample. The findings are in agreement with the reports by L’Arrive (1963), Bailey (1968) and Fingler (1982) who found that there were no significant differences in the accuracy of colony infestation as determined by the two methods. Thus either method can be used for routine spring examinations for Nosema incidence in bee colonies.

The differences in the magnitude of infestation between the samples of living bees and dead bees originating from the same colonies and the absence of unequivocal correlation between the two sets of samples may be an indication that those two kinds of samples do not reflect to the same extent the health status of wintering bees. The number of spores determined in the ground suspensions of dead bees was mostly higher than that found in preparations made from living bees. In the dead bee samples the percentage of infested worker bees was also higher. It can be assumed that in the whole number of dead bees those which perished because of parasite multiplication constituted a substantial percentage. It is noteworthy that in this study highly significant, positive correlation between nosema infestation of living bees and dead bees occurred only in December and January regardless of the method employed. Thus it was observed in the period in which Nosema incidence in wintering colonies is not high yet. It seems that at that time the multiplication of the protozoan is too weak to bring about high mortality rate of worker bees. Consequently, at that time diagnostic differences between the estimated levels of Nosema infestation of dead and living bee samples may be small and mutually correlated. The occurrence of the spring infection peak changes the general image of the infestation in the two kinds of samples. It also increases the percentage of heavily infested bees in the overall number of dead bees. At the same time, during the cleaning flight a certain number of spores is removed with the faeces of worker bees surviving in the cluster and, consequently, the infestation level is decreased.

The findings of this study bear out the suggestions of Fries et al. (1984) concerning the lesser suitability of dead bees from the so-called winter fall for the estimation of Nosema disease of honeybee colonies.

strona 137

REFERENCES

Bailey L. (1953) – The transmission of nosema disease. Bee Wld., 34(9):171-172.

Bailey L. (1967) – Nosema apis and dysentery of the honey bee. J.Apic.Res., 6(3):121-125.

Bailey L. (1968) – The measurement and interrelationships of infections with Nosema apis Malpighamoeba mellificae of honey bee populations. J.Invert.Path., 12:175-179.

Cantwell G.E. (1970) – Standard methods for counting nosema spores. Am.Bee J., 100(6):222-223.

Doull K.M. (1965) – The effects of time of day and method of sampling on the determination of nosema disease in beehives. J.Invert.Path., 7:1-14.

Doull K.M., Cellier K.M. (1961) – A survey of incidence of nosema disease (Nosema apis Zander) of the honey bees in South Australia. J.Insect Path., 3:280-288.

Doull K.M., Eckert J.E. (1962) – A survey of the incidence of nosema disease in California. J.Econ.Ent., 55:313-317.

Fingler B.G., Nash W.T., Szabo T.I. (1982) – A comparison of two techniques for the measurement of nosema disease in honeybee colonies wintered in Alberta, Canada. Am.Bee J., 122:319-361.

Fries I., Ekbohm G., Villumstad E. (1984) – Nosema apis, sampling techniques and honey yield. J.Apic.Res., 23:102-105.

Gochnauer T.A. (1957) – Fumidil B. and nosema control in wintering colonies. Glean. Bee Cult., 85:528-529.

Gross K.P., Ruttner F. (1970) – Enwickelt Nosema apis Zander eine Resistenz gegenüber dem Antibioticum Fumidil B? Apidologie, 1(4):401-421.

L’Arrivee J.C.M. (1963) – The effects of sampling sites on nosema determination J. Insect Path., 5:349-355.

Laere O. van, Weghe L. van de, Moermans R., Fordeyn D. (1980) – Quantitative determination of Nosema apis Z. in the individual honeybee, Apis mellifera L. Parasitica, 36:68-73.

Muszyńska J., Bornus L. (1983a) – Studies on the physiology of wintering in bee families. I. Assessment of wintering success in bee families from observation of their winter fall. Pszczeln.Zesz.Nauk., 27:3-21.

Muszyńska J., Bornus L. (1983b) – Studies on the physiology of wintering in bee families. II. A comparison of surviving bees from the cluster and dying from the winter fall. Pszczeln.Zesz.Nauk., 27:23-31.

Pickard R.S., El-Shemy A.A.M. (1989) – Seasonal variation in the infection of honeybee colonies with Nosema apis Zander. J.Apic.Res., 28(2):93-100.

PORÓWNANIE RÓŻNYCH METOD SŁUŻĄCYCH DO OCENY POZIOMU PORAŻENIA SPORAMI NOSEMA APIS Z. ZIMUJĄCYCH RODZIN PSZCZELICH

K. Pohorecka, J. Muszyńska

Streszczenie

Diagnozowanie nosemozy w rodzinach pszczelich oparte jest na stwierdzeniu metodami laboratoryjnymi obecności spor w tkankach i treści przewodu pokarmowego robotnic. Istnieje przy tym duże zróżnicowanie technik stosowanych do określania nasilenia inwazji Nosema apis w rodzinach pszczelich. Dotyczy ono nie tylko rodzaju próbek pszczół

strona 138

pobieranych do badań, ale także metody przeprowadzania oznaczeń mikroskopowych, a w konsekwencji sposobu przedstawiania wyników.

Do badań pobierane są zarówno pszczoły żywe jak i martwe (z tzw. osypu) w liczbie od 10 do 100 osobników. Pszczoły z danej próby mogą być badane indywidualnie (wynik badań najczęściej wyrażony procentem zarażonych pszczół w próbie) lub łącznie, tworząc próbę zbiorczą (wynik wyrażony średnią liczbą spor/pszczołę).

Badania, których wyniki przedstawiono w niniejszej pracy, miały dwa zasadnicze cele:

  1. Ocenę zależności pomiędzy dwoma metodami określania stopnia porażenia sporami N.apis zimujących rodzin pszczelich:
  2. metoda I – na podstawie średniej liczby spor/ pszczołę
  3. metoda II – na podstawie procentu porażonych pszczół w próbie
  4. Ocenę zależności pomiędzy dwoma rodzajami prób pszczół służącymi do określania stopnia porażenia sporami N.apis zimujących rodzin pszczelich:
  5. próby pszczół żywych (z kłębu)
  6. próby pszczół martwych (z osypu)

W każdym z trzech lat badaniami objęto po 6 rodzin pszczelich. Próby pszczół do badań w kierunku nosemozy pobierane były w dwóch kolejnych latach od listopada do kwietnia, w ostatnim roku do marca, w odstępach comiesięcznych. W tym samym terminie z każdej rodziny pobierane były do analiz jednocześnie dwa typy pszczół:

  • żywe – pobierane ze skraju kłębu bez jego naruszania,
  • martwe – które spadły na wkładkę podramkową w ciągu 7 dni poprzedzających termin pobrania próby (Muszyńska, Bornus 1983a). Każda próba liczyła 10 pszczół.

Ocenę porażenia pszczół w próbie prowadzono w oparciu o badania indywidualnych robotnic przy użyciu hemacytometru Bürkera. Wyniki uzyskane dla każdej z 10 pszczół sumowano, a następnie wyliczano średnią liczbę spor przypadającą na jedną pszczołę w próbie (metoda I) oraz, jaki procent próby stanowią porażone pszczoły (metoda II). Wyniki badań opracowano statystycznie metodą analizy regresji.

We wszystkich przeprowadzonych analizach stwierdzono istnienie dodatnich, wysoce istotnych korelacji pomiędzy wynikami porażenia sporami N.apis rodzin pszczelich, wyrażonymi średnią liczbą spor/pszczołę w próbie a wynikami porażenia wyrażonymi procentem zarażonych pszczół w próbie. Współzależności te wystąpiły zarówno przy ocenie wyników odnoszących się do prób pszczół żywych pochodzących z kłębu jak i prób pszczół martwych pochodzących z osypu.

Uzyskane wyniki sugerują, że średnia liczba spor Nosema apis przypadających na 1 pszczołę w próbie liczącej 10 osobników pozostaje w ścisłym związku z procentem zakażonych pszczół w danej próbie. Zależności te występują niezależnie od rodzaju pszczół jakie pobierane były do oznaczeń. Oznacza to, iż dokładność obu metod oznaczania wielkości porażenia sporami N.apis jest porównywalna.

Natomiast porównanie średniej liczby spor przypadającej na pszczołę żywą ze średnią liczbą spor przypadającą na pszczołę martwą w poszczególnych próbach pochodzących z tych samych rodzin wykazało brak istnienia pomiędzy nimi współzależności, w większości wykonanych analiz.

Analizy oparte na porównaniu procentu zarażonych robotnic w próbach pszczół pobranych z kłębu, z procentem zarażonych robotnic w próbach pszczół pochodzących z osypu również nie dały jednoznacznych wyników.

Uzyskane wyniki świadczą o istnieniu rozbieżności między wielkością porażenia sporami N.apis rodzin pszczelich, obliczoną na podstawie badania próbek pszczół żywych pobranych z kłębu a wielkością porażenia uzyskaną na drodze badania próbek pszczół

strona 139

martwych pochodzących z osypu. Różnice te pojawiły się niezależnie od stosowanej metody oceny porażenia, a poziom infekcji był zawsze wyższy w próbkach pszczół martwych.

Potwierdza to sugestie Friesa i innych (1984), o mniejszej przydatności pszczół martwych z tzw. osypu zimowego do określenia faktycznego nasilenia nosemozy wiosną w rodzinach pszczelich.

Słowa kluczowe: Nosema apis, stopień porażenia, diagnostyka, metoda.


Errata — błędy druku znalezione podczas digitalizacji

  • s. 129, streszczenie: „measurre” → measure
  • s. 130, 3×: „Doul” → Doull
  • s. 130: „Celier” → Cellier
  • s. 130: „level that the” → level than the
  • s. 130: „L’Arrivae” → L’Arrivee
  • s. 133: „estimated two methods” → estimated by two methods
  • s. 136: „L’Arrive (1963)” → L’Arrivee (1963)
  • s. 137, piśmiennictwo: „apis Malpighamoeba” → apis and Malpighamoeba
  • s. 137, piśmiennictwo: „Enwickelt” → Entwickelt

Podobne wpisy