As pastagens são o principal recurso alimentar para pecuária brasileira, sendo fundamentais na viabilização dos sistemas produtivos. Fatores climáticos como temperatura e fotoperíodo têm grande importância no crescimento das plantas forrageiras, influenciando o acúmulo de forragem a distribuição estacional da produção e do valor nutritivo, fatores de grande impacto no sistema produtivo como um todo. Dessa forma, o desenvolvimento de ferramentas que racionalizem o processo produtivo das plantas possibilitando a sua predição em função dos parâmetros ambientais que o condicionam são importantes para o planejamento desses sistemas. O presente estudo foi desenvolvido no Departamento de Zootecnia da ESALQ/USP em Piracicaba, SP, de Dezembro de 2000 a Março de 2002, com o objetivo de gerar um banco de dados sobre as características produtivas e qualitativas de gramíneas forrageiras do gênero Cynodon e, a partir desse banco de dados, avaliar modelos matemáticos propostos na literatura para estimar o acúmulo de massa seca e as alterações no valor nutritivo [teores de proteína bruta (PB), fibra em detergente neutro (FDN), e a digestibilidade in vitro da matéria orgânica (DIVMO)] dessas forragens. Foram estudadas duas variáveis climáticas, uma gerada em função da soma térmica diária (graus-dia, GD), e outra incorporando os efeitos das alterações no fotoperíodo à soma térmica (a unidade fototérmica, UF). Os tratamentos constituiram de todas as combinações possíveis entre dois intervalos entre cortes (4 e 6 semanas) e cinco cultivares de Cynodon spp. (Tifton 85, Coastcross, Florico, Florona e Estrela) mecanicamente colhidos a 7 cm. As parcelas eram irrigadas para garantir ausência de déficit hídrico e adubadas com o equivalente a 400 kg N ha-1 ano-1. O delineamento experimental utilizado foi o de blocos completos casualizados com quatro repetições. Nas três estações de crescimento determinadas, "estação completa", "verão" e "inverno" foram avaliados o acúmulo de forragem (AF), a taxa média diária de acúmulo de forragem (TMDAF), a distribuição estacional da produção e os teores de PB e de FDN, além da DIVMO. O intervalo de seis semanas proporcionou maior AF, maior TMDAF, menor teor de PB, menor DIVMO e maior teor de FDN do que o intervalo de quatro semanas nas três "estações" consideradas. A distribuição estacional da produção foi igual para os dois intervalos, com 73% da produção no "verão" e 27% no "inverno". O AF e a TMDAF para a "estação completa" não foram diferentes entre os cultivares, mas no "verão" Tifton 85 foi superior a Coastcross e Florona. No "inverno", os cultivares não se diferenciaram nessas características, o que resultou em maior estacionalidade de produção de Tifton 85 do que de Coastcross, Florico e Florona. Coastcross apresentou os menores teores de PB e juntamente com Tifton 85 os maiores teores de FDN nas três"estações". A digestibilidade só foi diferente entre os cultivares na "estação completa", onde Florona superou Coastcross. Os modelos gerados para a resposta quantitativa (AF) foram altamente significativos, apresentando melhores resultados nos menores níveis de agregação. As duas variáveis climáticas foram eficientes, mas os modelos de UF apresentaram capacidade preditória superior. Os modelos para as respostas qualitativas também apresentaram alta significância, com melhores resultados nos maiores níveis de agregação. No entanto, em função da alta variabilidade tiveram sua capacidade preditória comprometida. Os cultivares estudados demonstraram significativas diferenças produtivas e qualitativas, que devem ser consideradas quando de sua adoção em sistemas produtivos. Nos intervalos entre cortes a produção e o valor nutritivo estiveram inversamente associados. A utilização de variáveis climáticas para a predição de parâmetros produtivos se mostrou efetiva em condições de ausência de déficit hídrico. Para a modelagem qualitativa o conceito se mostrou válido, mas novos modelos incluindo maior amplitude de maturidades da forragem devem ser gerados e avaliados. / Pastures are the main feed resource in the Brazilian livestock industry and they are key in making forage-livestock systems feasible. Climatic variables such as temperature and daylength are important to forage growth as they affect herbage accumulation as well as the seasonal distribution of both yield and nutritive value, two major characteristics that impact the systems as a whole. Thus, the development of manegerial tools that allow for the rationalization of the production process and for the prediction of forage responses to environmental variables may be valuable for planning and managing whole systems. The present study was carried out at the Departamento de Zootecnia of ESALQ-USP in Piracicaba, SP, from December 2000 through March 2002. The objective was to generate a comprehensive dataset on the productive and qualitative characteristics of Cynodon grasses and to use this dataset to evaluate prediction models reported in the literature, using them to estimate forage accumulation and changes in forage nutritive value [measured as concentrations of crude protein (CP) and neutral detergent fiber (NDF), as well as the in vitro organic matter digestibility (IVOMD)]. Two climatic variables were studied, one generated as a function of daily caloric sum (degree-days, DD) and the other combining the effect of daylength with DD and known as the photothermal unit (PU). Treatments included all possible combinations between two intervals between clippings (every four or six weeks) and five Cynodon spp. cultivars (Tifton 85, Coastcross, Florico, Florona e Stargrass) harvested mechanically at a 7-cm height. Plots were irrigated to ensure that soil moisture was not limiting at any point during the experiment and fertilized at the rate of 400 kg N ha-1 yr-1. The experimental design was a randomized complete block with four replications. For the three "seasons" established ("whole season", "summer", and "winter"), total herbage accumulation (HA), the mean daily HA rate (MDHAR), the seasonal yield distribution and the concentrations of CP and NDF plus the IVOMD of the forage produced were characterized. The 6-wk harvest interval resulted in higher HA and MDHAR, lower CP and IVOMD, and higher NDF than the 4-wk interval, for all three "seasons". Seasonal yield distribution was similar between intervals, with 73% of the total forage accumulating during the "summer" and 27% in the "winter". HA and MDHAR for the "whole season" was the same across cultivars but during the "summer", these responses were higher in Tifton 85 than on Coastcross and Stargrass. In the "winter", no differences were found across cultivars for these quantitative responses and this resulted in more pronounced seasonality for Tifton 85 than for Coastcross, Florico and Florona. Coastcross forage had the lowest CP concentration and, together with Tifton 85 forage, the highest NDF concentrations in all three "seasons". Forage IVOMD was the same across treatments except among cultivars in the "whole season", where Florona showed higher IVOMD than Coastcross. The prediction models generated for the quantitative response HA were highly significant and seem to be better predictors at lower aggregation levels. Both climatic variables were efficient, although the PU models had a better prediction ability than the DD models. For the qualitative characteristics were also significant, but with better results for the higher aggregation levels. Because of the high variability involved, however, their predictive ability may have been somewhat compromised. The grass cultivars used in this study have contrasting productive and qualitative characteristics, which must be taken into account when one considers their inclusion into the system. Within harvest intervals yield and nutritive value were inversely related. The use of climatic variables to predict forage quantitative responses appears to be promising in the absence of water deficit. For qualitative characteristics, the concept seems to hold although more modeling is needed using a wider range of forage maturities.

Ptilotus nobilis is a short-lived perennial wildflower, native to semi-arid and arid areas of Australia. Propagation by vegetative means is constrained by the early onset and a continuous flowering habit. Despite being defined as the main barrier for vegetative propagation no published research on floral induction and initiation was found. The aim of this study is to provide the first insights into floral evocation in P. nobilis, with the general objectives being to investigate floral induction and initiation, find the means to maintain plants in the vegetative phase, enhance leaf initiation and to gain knowledge of growth and development with an emphasis on light intensity, temperature and daylength. Early experiments aimed to increase the understanding of growth and development. The main cardinal events were identified including the onset of branching (axillary stem growth), visible bud stage, first floret opening and maturity of the inflorescence. To aid future cultivation schedules, four growing degree days (GDD) and one chronological model, using days after sowing (DAS), were established and validated. All GDD models were accurate in predicting first floret opening and maturity, but not in predicting visible bud stage. Best prediction was achieved by using an upper temperature threshold of 18.2 ºC and an own base temperature of 5.0 ºC. Days after sowing were an accurate means of prediction, indicating that temperature and other variables such as light intensity regulate development. A series of defoliation treatments investigated the maintenance of P. nobilis plants in the vegetative phase. Within all treatments, reproductive structures were observed and all plants with more than two true leaves entered the reproductive phase. Plants with less than two true leaves showed a delayed floral bud appearance by up to 20 days. Floral development was affected by most defoliation treatments resulting in vegetative growth within the inflorescences. It is hypothesized that plants have a very short juvenile phase and that a constant floral stimulus may be needed for floral evocation. Using scanning electron and light microscopy a template for the transition from vegetative to reproductive phase was developed. Bract initiation was accompanied by a significant increase in meristem area and diameter, and was defined as the onset of flowering. The established template was used in a subsequent glasshouse trial, which revealed that floral initiation occurred very early and at 25 DAS all plants had entered the reproductive phase. Branching and leaf area expansion were identified as post- initiation processes. Plants were exposed to different light intensities (229.3, 398.6 and 909.3 µmol m-2 s-1) in a glasshouse. Under low light, final leaf number increased by up to three leaves, indicating that the vegetative phase was prolonged. Cardinal events were delayed but all plants reached maturity. It was concluded that a light intensity of 229.3 µmol m-2 s-1 was not low enough to prevent floral initiation and that a further decrease of light intensity in combination with temperature could be more effective. Interactions of light intensity, temperature and daylength were investigated. Plants were grown under high light and low light (< 1.2 MJ m-2) intensities, 25/10 ºC and 35/20 ºC and daylengths of 11 h and 16 h. During the trial period (42 days), low light intensity suppressed floral initiation and high temperatures maintained more plants (70%) in the vegetative phase. However daylength treatments had no effect on the time of bract initiation or the percentage of vegetative plants. After 83 days floral buds and axillary stems were observed on some plants under low light intensity, indicating the onset of the reproductive phase and showing that P. nobilis could not be maintained in the vegetative phase indefinitely. Plants under high light were harvested at maturity and effects of temperature and daylength were analysed. Final leaf number increased under 35/20 ºC supporting the previously established results that floral initiation was delayed and leaf production enhanced by high temperature. Morphological data was collected to classify the photoperiodic response of P. nobilis. Plants under 25/10 ºC and 11 h had the longest inflorescences and greatest number of buds and flowers, which was also reflected in the buds and flowers dry weight, with an increase of up to 3.4 fold under these conditions. The difficulty of classifying some plants accordingly to their photoperiodic response and the proposal that P. nobilis may be a facultative short day plant under 25/10 ºC is discussed. In summary, this study presents the first evidence that P. nobilis has a very short juvenile phase and that growth and development are mainly driven by light intensity and temperature. It was possible to enhance vegetative growth by defoliation, low light intensity and high temperatures, however this did not fully prevent flowering, which indicates that P. nobilis has a very strong flowering response or signal.

Ptilotus nobilis is a short-lived perennial wildflower, native to semi-arid and arid areas of Australia. Propagation by vegetative means is constrained by the early onset and a continuous flowering habit. Despite being defined as the main barrier for vegetative propagation no published research on floral induction and initiation was found. The aim of this study is to provide the first insights into floral evocation in P. nobilis, with the general objectives being to investigate floral induction and initiation, find the means to maintain plants in the vegetative phase, enhance leaf initiation and to gain knowledge of growth and development with an emphasis on light intensity, temperature and daylength. Early experiments aimed to increase the understanding of growth and development. The main cardinal events were identified including the onset of branching (axillary stem growth), visible bud stage, first floret opening and maturity of the inflorescence. To aid future cultivation schedules, four growing degree days (GDD) and one chronological model, using days after sowing (DAS), were established and validated. All GDD models were accurate in predicting first floret opening and maturity, but not in predicting visible bud stage. Best prediction was achieved by using an upper temperature threshold of 18.2 ºC and an own base temperature of 5.0 ºC. Days after sowing were an accurate means of prediction, indicating that temperature and other variables such as light intensity regulate development. A series of defoliation treatments investigated the maintenance of P. nobilis plants in the vegetative phase. Within all treatments, reproductive structures were observed and all plants with more than two true leaves entered the reproductive phase. Plants with less than two true leaves showed a delayed floral bud appearance by up to 20 days. Floral development was affected by most defoliation treatments resulting in vegetative growth within the inflorescences. It is hypothesized that plants have a very short juvenile phase and that a constant floral stimulus may be needed for floral evocation. Using scanning electron and light microscopy a template for the transition from vegetative to reproductive phase was developed. Bract initiation was accompanied by a significant increase in meristem area and diameter, and was defined as the onset of flowering. The established template was used in a subsequent glasshouse trial, which revealed that floral initiation occurred very early and at 25 DAS all plants had entered the reproductive phase. Branching and leaf area expansion were identified as post- initiation processes. Plants were exposed to different light intensities (229.3, 398.6 and 909.3 µmol m-2 s-1) in a glasshouse. Under low light, final leaf number increased by up to three leaves, indicating that the vegetative phase was prolonged. Cardinal events were delayed but all plants reached maturity. It was concluded that a light intensity of 229.3 µmol m-2 s-1 was not low enough to prevent floral initiation and that a further decrease of light intensity in combination with temperature could be more effective. Interactions of light intensity, temperature and daylength were investigated. Plants were grown under high light and low light (< 1.2 MJ m-2) intensities, 25/10 ºC and 35/20 ºC and daylengths of 11 h and 16 h. During the trial period (42 days), low light intensity suppressed floral initiation and high temperatures maintained more plants (70%) in the vegetative phase. However daylength treatments had no effect on the time of bract initiation or the percentage of vegetative plants. After 83 days floral buds and axillary stems were observed on some plants under low light intensity, indicating the onset of the reproductive phase and showing that P. nobilis could not be maintained in the vegetative phase indefinitely. Plants under high light were harvested at maturity and effects of temperature and daylength were analysed. Final leaf number increased under 35/20 ºC supporting the previously established results that floral initiation was delayed and leaf production enhanced by high temperature. Morphological data was collected to classify the photoperiodic response of P. nobilis. Plants under 25/10 ºC and 11 h had the longest inflorescences and greatest number of buds and flowers, which was also reflected in the buds and flowers dry weight, with an increase of up to 3.4 fold under these conditions. The difficulty of classifying some plants accordingly to their photoperiodic response and the proposal that P. nobilis may be a facultative short day plant under 25/10 ºC is discussed. In summary, this study presents the first evidence that P. nobilis has a very short juvenile phase and that growth and development are mainly driven by light intensity and temperature. It was possible to enhance vegetative growth by defoliation, low light intensity and high temperatures, however this did not fully prevent flowering, which indicates that P. nobilis has a very strong flowering response or signal.

Abstract Adaptation to local environment is important for all organisms to guarantee survival and to maximize reproduction. Populations of the same species may live in environments that differ markedly. Due to differential selection pressures this can lead to population differentiation, which can be studied both at the phenotypic and at the gene level. The growing season cued by long days is typically short in the north, whereas southern populations have long growing seasons and are adapted to short days. Seasonal fluctuations in temperature also differ between northern and southern environments. Daylength and temperature regulate the timing of flowering in plants. Environmental regulation of flowering and its genetic basis has been extensively studied in the annual model species Arabidopsis thaliana. The perennial growth and flowering habit has been studied especially in trees, but studies on herbaceous plants species have been lacking. In this thesis, I have studied adaptation to growing season length in a perennial herbaceous model species Arabidopsis lyrata. Individuals from populations adapted to northern and southern environments in Europe were grown in same conditions in the growth chambers and in the field. Differentiation between the populations was studied by observing their flowering phenotypes and by studying the expression of genes that are candidates for governing the phenotypic differentiation. The main result in the thesis was that adaptation to short growing season in north can be seen as long daylength requirement for flowering and as fast developmental rate. Critical daylength for flowering likely regulates especially the timing of flowering cessation in the end of the growing season. Flowering time of individuals from northern populations also responded more strongly to cold treatment (representing winter) than that of the southern population. The cold requirement for flowering guarantees that the plants only flower after the winter in the spring with suitable conditions. Expression studies indicated that population differentiation in flowering could be at least partly governed by the expression variation in a few candidate genes. The results in this thesis are valuable for instance for understanding perennial species in general, including tree and crop species, and for predicting how plants response to changing climate. / Tiivistelmä Ympäröiviin oloihin sopeutuminen on tärkeää kaikille organismeille selviytymisen ja jälkeläistuoton kannalta. Saman lajin eri populaatiot saattavat elää ympäristöissä, joiden olosuhteet poikkeavat toisistaan huomattavasti. Tällöin populaatioihin kohdistuvat erilaiset valintapaineet ja populaatiot erilaistuvat. Erilaistuminen havaitaan tarkastelemalla yksilöiden ilmiasuja ja geenejä. Kasvukauden pituus määrittää eteläisten ja pohjoisten kasvuympäristöjen valintapaineita. Pohjoisessa kasvukausi on lyhyt ja sen alkamisesta ja loppumisesta kertoo pitkä päivänpituus. Etelässä on pitkä kasvukausi ja siellä elävät populaatiot ovat sopeutuneet lyhyeen päivänpituuteen. Myös vuodenaikaiset lämpötilavaihtelut eroavat pohjoisten ja eteläisten alueiden välillä. Muutokset päivänpituudessa ja lämpötilassa säätelevät kasveilla kukkimisen ajankohtaa. Kukkimiseen vaikuttavia ympäristötekijöitä ja kukkimista sääteleviä geenejä on tutkittu paljon yksivuotisella mallilajilla lituruoholla (Arabidopsis thaliana). Monivuotisten kasvien kasvun ja kukkimisen säätelyä on tutkittu etenkin puilla, mutta hyvin vähän ruohovartisilla kasveilla. Tässä väitöskirjatyössä tutkin kasvukauden pituuteen sopeutumista monivuotisella ruohovartisella lajilla, idänpitkäpalolla (Arabidopsis lyrata). Pohjoisiin ja eteläisiin ympäristöoloihin sopeutuneiden eurooppalaisten populaatioiden yksilöitä kasvatettiin samanlaisissa olosuhteissa kontrolloiduissa kasvatushuoneissa ja kenttäolosuhteissa. Populaatioiden erilaistumista tarkkailtiin kukkimiseen liittyvissä ominaisuuksissa sekä eroja selittävien kandidaattigeenien ekspressiossa. Tutkimuksen päätulos oli, että pohjoisen populaation lyhyeen kasvukauteen sopeutuminen voidaan havaita pitkän päivän vaatimuksena kukkimiselle ja nopeana kehityksenä. Kenttäolosuhteissa päivänpituus sääteli etenkin kukkimisen lopetusta kasvukauden lopussa. Pitkä kylmäkäsittely nopeutti kukkimista etenkin pohjoisessa populaatiossa. Kukkimista edeltävä kylmävaatimus takaa, etteivät kasvit kuki syksyllä epäsuotuisissa olosuhteissa, vaan vasta keväällä talven jälkeen. Populaatioiden erilaistuminen kukkimisen päivänpituus- ja kylmävaatimuksessa selittyy todennäköisesti ainakin osittain kandidaattigeenien ekspressioeroilla. Tämän väitöskirjatyön tuloksia voidaan soveltaa monivuotisten viljelykasvien jalostuksessa tai ennustettaessa kasvipopulaatioiden sopeutumista ilmastonmuutokseen.

As pastagens são o principal recurso alimentar para pecuária brasileira, sendo fundamentais na viabilização dos sistemas produtivos. Fatores climáticos como temperatura e fotoperíodo têm grande importância no crescimento das plantas forrageiras, influenciando o acúmulo de forragem a distribuição estacional da produção e do valor nutritivo, fatores de grande impacto no sistema produtivo como um todo. Dessa forma, o desenvolvimento de ferramentas que racionalizem o processo produtivo das plantas possibilitando a sua predição em função dos parâmetros ambientais que o condicionam são importantes para o planejamento desses sistemas. O presente estudo foi desenvolvido no Departamento de Zootecnia da ESALQ/USP em Piracicaba, SP, de Dezembro de 2000 a Março de 2002, com o objetivo de gerar um banco de dados sobre as características produtivas e qualitativas de gramíneas forrageiras do gênero Cynodon e, a partir desse banco de dados, avaliar modelos matemáticos propostos na literatura para estimar o acúmulo de massa seca e as alterações no valor nutritivo [teores de proteína bruta (PB), fibra em detergente neutro (FDN), e a digestibilidade in vitro da matéria orgânica (DIVMO)] dessas forragens. Foram estudadas duas variáveis climáticas, uma gerada em função da soma térmica diária (graus-dia, GD), e outra incorporando os efeitos das alterações no fotoperíodo à soma térmica (a unidade fototérmica, UF). Os tratamentos constituiram de todas as combinações possíveis entre dois intervalos entre cortes (4 e 6 semanas) e cinco cultivares de Cynodon spp. (Tifton 85, Coastcross, Florico, Florona e Estrela) mecanicamente colhidos a 7 cm. As parcelas eram irrigadas para garantir ausência de déficit hídrico e adubadas com o equivalente a 400 kg N ha-1 ano-1. O delineamento experimental utilizado foi o de blocos completos casualizados com quatro repetições. Nas três estações de crescimento determinadas, &#34;estação completa&#34;, &#34;verão&#34; e &#34;inverno&#34; foram avaliados o acúmulo de forragem (AF), a taxa média diária de acúmulo de forragem (TMDAF), a distribuição estacional da produção e os teores de PB e de FDN, além da DIVMO. O intervalo de seis semanas proporcionou maior AF, maior TMDAF, menor teor de PB, menor DIVMO e maior teor de FDN do que o intervalo de quatro semanas nas três &#34;estações&#34; consideradas. A distribuição estacional da produção foi igual para os dois intervalos, com 73% da produção no &#34;verão&#34; e 27% no &#34;inverno&#34;. O AF e a TMDAF para a &#34;estação completa&#34; não foram diferentes entre os cultivares, mas no &#34;verão&#34; Tifton 85 foi superior a Coastcross e Florona. No &#34;inverno&#34;, os cultivares não se diferenciaram nessas características, o que resultou em maior estacionalidade de produção de Tifton 85 do que de Coastcross, Florico e Florona. Coastcross apresentou os menores teores de PB e juntamente com Tifton 85 os maiores teores de FDN nas três&#34;estações&#34;. A digestibilidade só foi diferente entre os cultivares na &#34;estação completa&#34;, onde Florona superou Coastcross. Os modelos gerados para a resposta quantitativa (AF) foram altamente significativos, apresentando melhores resultados nos menores níveis de agregação. As duas variáveis climáticas foram eficientes, mas os modelos de UF apresentaram capacidade preditória superior. Os modelos para as respostas qualitativas também apresentaram alta significância, com melhores resultados nos maiores níveis de agregação. No entanto, em função da alta variabilidade tiveram sua capacidade preditória comprometida. Os cultivares estudados demonstraram significativas diferenças produtivas e qualitativas, que devem ser consideradas quando de sua adoção em sistemas produtivos. Nos intervalos entre cortes a produção e o valor nutritivo estiveram inversamente associados. A utilização de variáveis climáticas para a predição de parâmetros produtivos se mostrou efetiva em condições de ausência de déficit hídrico. Para a modelagem qualitativa o conceito se mostrou válido, mas novos modelos incluindo maior amplitude de maturidades da forragem devem ser gerados e avaliados. / Pastures are the main feed resource in the Brazilian livestock industry and they are key in making forage-livestock systems feasible. Climatic variables such as temperature and daylength are important to forage growth as they affect herbage accumulation as well as the seasonal distribution of both yield and nutritive value, two major characteristics that impact the systems as a whole. Thus, the development of manegerial tools that allow for the rationalization of the production process and for the prediction of forage responses to environmental variables may be valuable for planning and managing whole systems. The present study was carried out at the Departamento de Zootecnia of ESALQ-USP in Piracicaba, SP, from December 2000 through March 2002. The objective was to generate a comprehensive dataset on the productive and qualitative characteristics of Cynodon grasses and to use this dataset to evaluate prediction models reported in the literature, using them to estimate forage accumulation and changes in forage nutritive value [measured as concentrations of crude protein (CP) and neutral detergent fiber (NDF), as well as the in vitro organic matter digestibility (IVOMD)]. Two climatic variables were studied, one generated as a function of daily caloric sum (degree-days, DD) and the other combining the effect of daylength with DD and known as the photothermal unit (PU). Treatments included all possible combinations between two intervals between clippings (every four or six weeks) and five Cynodon spp. cultivars (Tifton 85, Coastcross, Florico, Florona e Stargrass) harvested mechanically at a 7-cm height. Plots were irrigated to ensure that soil moisture was not limiting at any point during the experiment and fertilized at the rate of 400 kg N ha-1 yr-1. The experimental design was a randomized complete block with four replications. For the three &#34;seasons&#34; established (&#34;whole season&#34;, &#34;summer&#34;, and &#34;winter&#34;), total herbage accumulation (HA), the mean daily HA rate (MDHAR), the seasonal yield distribution and the concentrations of CP and NDF plus the IVOMD of the forage produced were characterized. The 6-wk harvest interval resulted in higher HA and MDHAR, lower CP and IVOMD, and higher NDF than the 4-wk interval, for all three &#34;seasons&#34;. Seasonal yield distribution was similar between intervals, with 73% of the total forage accumulating during the &#34;summer&#34; and 27% in the &#34;winter&#34;. HA and MDHAR for the &#34;whole season&#34; was the same across cultivars but during the &#34;summer&#34;, these responses were higher in Tifton 85 than on Coastcross and Stargrass. In the &#34;winter&#34;, no differences were found across cultivars for these quantitative responses and this resulted in more pronounced seasonality for Tifton 85 than for Coastcross, Florico and Florona. Coastcross forage had the lowest CP concentration and, together with Tifton 85 forage, the highest NDF concentrations in all three &#34;seasons&#34;. Forage IVOMD was the same across treatments except among cultivars in the &#34;whole season&#34;, where Florona showed higher IVOMD than Coastcross. The prediction models generated for the quantitative response HA were highly significant and seem to be better predictors at lower aggregation levels. Both climatic variables were efficient, although the PU models had a better prediction ability than the DD models. For the qualitative characteristics were also significant, but with better results for the higher aggregation levels. Because of the high variability involved, however, their predictive ability may have been somewhat compromised. The grass cultivars used in this study have contrasting productive and qualitative characteristics, which must be taken into account when one considers their inclusion into the system. Within harvest intervals yield and nutritive value were inversely related. The use of climatic variables to predict forage quantitative responses appears to be promising in the absence of water deficit. For qualitative characteristics, the concept seems to hold although more modeling is needed using a wider range of forage maturities.